Descarregar

Transcription

Descarregar
Phenomenology of Non-Standard Neutrino
Interactions
Francisco Javier Escrihuela Ferrándiz
Departament de Fı́sica Teòrica
Tesi Doctoral
Director: Prof. José Wagner Furtado Valle
Co-Director: Prof. Omar Gustavo Miranda Romagnoli
Co-Directora: Dra. Marı́a Amparo Tórtola Baixauli
2016
2
Certificat
Prof. José W. Furtado Valle, professor d’investigació del IFIC, centre mixt del CSIC
i de la Universitat de València,
Dra. Marı́a Amparo Tórtola Baixauli, investigadora Ramón y Cajal de la Universitat
de València,
Prof. Omar Gustavo Miranda Romagnoli, professor d’investigació del CINVESTAV
de Mèxic D.F.,
CERTIFIQUEN:
Que la present memòria, “Phenomenology of Non-Standard Neutrino Interaccions”,
ha estat realitzada sota la seva direcció al Departament de Fı́sica Teòrica de la Universitat de València per Francisco Javier Escrihuela Ferrándiz, i que constitueix la seva tesi
doctoral per optar al grau de Doctor en Fı́sica.
Perquè aixı́ conste, en compliment de la legislació vigent, presenten al Departament de
Fı́sica Teòrica de la Universitat de València la referida Tesi doctoral, i signen el present
certificat.
Prof. José W. Furtado Valle
Dra. Marı́a Amparo Tórtola Baixauli
Prof. Omar Gustavo Miranda Romagnoli
3
4
Agraı̈ments
M’agradaria expressar la més sincera gratitud a tots aquells que han fet possible la realització d’aquesta tesi.
En primer lloc a la meva famı́lia cientı́fica composada principalment pels meus directors
de tesi: José W. Furtado Valle, Marı́a Amparo Tórtola Baixauli i Omar G. Miranda
Romagnoli, vosaltres sabeu ben bé que aquest treball no haguera eixit mai endavant
sense vosaltres, i es “quasi” més vostre que meu. Gràcies José per haver-me donat la
oportunitat de treballar amb tu en les condicions tan especials que m’envoltaven; sempre
m’he sentit recolzat i m’has animat a seguir endavant. Gràcies Mariam perquè sempre
has estat ahı́, disposada a ajudar-me amb paciència i dedicació, sempre atenta a les meves
necessitats. Gràcies Omar pel teu tracte personal; m’has fet sentir com part de la famı́lia
i has sigut peça fonamental en la consecució d’aquesta tesi, bé que ho saps. De tots tres
m’emporte una nova forma de vore la Fı́sica i una estima incondicional pels neutrins.
Dins d’aquesta famı́lia cientı́fica estan també tots els membres del AHEP, encapçalats
per Martin i Sergio; sempre m’heu fet sentir com a casa. Voldria agrair també l’hospitalitat
que sempre m’ha brindat el CINVESTAV de Mèxic D.F, en especial a Àlex, Félix, Estela,
Ricardo i Blanca; i a la meva famı́lia mexicana, Luz, Gustavo i Lupita. Entre tots m’heu
fet els estius mexicans molt càlids i interessants.
Deixant a banda la Fı́sica m’agradaria agrair, per descomptat, a la meva famı́lia. Als
meus pares Paco i Maria, per donar-me una gran educació, suport incondicional i amor
durant tots els dies de la meva vida. Sempre esteu ahı́. Als meus germans Mariam i
Vı́ctor, perquè no seria el que sóc si no els haguera tingut sempre al meu costat. Moltes
gràcies als meus cunyats David i Ana, i als meus nebodets David i Lucı́a, per donar-li
calor i alegria a la meva vida.
Vull mostrar també el meu agraı̈ment als meus amics, perquè amb el vostre... “¡a ver
cuándo acabas!” o “tómate unas birras para despejarte”, m’heu donat forces per seguir
endavant. Gràcies a Pablo, Pepe B., José Alberto, Francis, Josemi, Jose, Ricard, Antonio,
Galera, Olaya, Pepe Ll., Joan, Gabi, Juan, Álex, Jorge, Jaumet, Abuelo... Espere no
5
6
deixar-me ningun... Gràcies també als companys de treball, que quan m’han vist aclaparat
m’han fet la vida més fàcil: Carles, J. Antonio, Maria, Marisa, Ródenas, H.Ana...
I l’últim però, això sı́, el més especial agraı̈ment és per a la meva dona Ana. La teva
paciència amb mi no té lı́mits. Sempre tens una paraula d’alè i un somriure a la cara. La
vida sense tu no tindria sentit.
Gràcies a tots! Aquesta tesi és un poc de tots vosaltres.
I want to thank Conacyt (Mexico) because part of this work has been supported by
Conacyt grant 166639.
Preface
Today neutrino physics is in a privileged position within the fascinating field of particle
physics. From the discovery of neutrino oscillations by Super-Kamiokande [1] in 1998, the
door to physics beyond the Standard Model (SM in what follows) has been opened. This
fact implies that neutrinos have to be massive [2] in opposition to the Standard Model
assumption. However, this is not a surprise completely, but it was already hinted from
theoretical and experimental observations in the two decades prior to the discovery of the
oscillatory phenomenon:
• In principle, it is not necessary to introduce neutrino masses in the Standard Model.
Nevertheless, this is not a characteristic arising from a gauge symmetry (as in the
photon case) but it is chosen for simplicity, avoiding, for instance, to introduce the
right-handed neutrino.
• Most of the unification models incorporate neutrino masses.
• The experimentally observed deficit of the atmospheric and solar neutrino fluxes
could be explained by the phenomenon of neutrino oscillations, mechanism that
requires neutrino mass and mixing.
As a consequence of this new path opened by neutrinos, one may observe processes
forbidden in the Standard Model [3]:
• Neutrino oscillations imply neutrino mixing; therefore the generational lepton numbers are not valid as a global symmetry. Decays with lepton flavor violation (LFV)
are possible, as for example:
– µ → eγ ,
– τ → eγ ,
– µ → 3e .
7
8
Preface
• If neutrinos are Majorana particles,neutrinoless double beta decay (0νββ) and other
processes with lepton number violation could occur. Until such historic observation
is made, the debate about the nature of neutrinos as Dirac or Majorana particles
will be open.
• Cosmology is sensitive to massive neutrinos, which may affect the cosmic microwave
background and other cosmological observables.
• The neutrino, as a massive particle, could decay:
– να → να 0 γ ,
– να → να0 να00 να000 , where three neutrinos are produced in the final state. These
could be the same kind of neutrino or not.
– να → να0 J , where J is a majoron 1 .
As we have already mentioned, none of the phenomena described in the above paragraphs
is foreseen in the SM. Therefore models beyond SM are required to introduce neutrino
mass and they may imply new interactions with matter: Non-Standard Interactions (NSI
in what follows). In this thesis we study some of these new interactions from a phenomenological point of view. To achieve it, we will use data from several neutrino experiments,
obtaining extra information on NSI stemming from a combined analysis of them. For this
purpose the thesis is organized as follows:
• Chapter 1: We begin with a review of the most important characteristics of neutrinos in the Standard Model. Along with them, we will also present other properties
of beyond the SM neutrinos, such as neutrino mass and neutrino oscillations.
• Chapter 2: In this chapter we will give an introduction to NSI, explaining their
appearance and main features. Along with this generic explanation, we will discuss
models where NSI appear spontaneously and their influence on neutrino oscillation
probabilities and experiments.
• Chapter 3: The main goal of this chapter is to show that, although playing a
secondary role, NSI still have an influence on the solution of the solar neutrino
problem. We will combine solar, reactor and accelerator data, in order to get limits
on parameters which determine the NSI strength.
1
The majoron J is a Goldstone boson which appears in models with spontaneous breaking of the
lepton number [4].
9
• Chapter 4: In this part of the thesis we will get limits on NSI parameters involved
in muon-neutrino interactions with quarks. For this purpose, we will use the results
from a NUTeV reanalysis performed by two collaborations, along with data coming
from accelerator and atmospheric neutrino experiments.
• Chapter 5: The non-unitarity of the light neutrino mixing matrix is the simplest
example of NSI. In this chapter we will describe a formalism which will allow us
to work easily with models where more than three neutrinos are considered (as
seesaw models), separating the new physics and the standard one. We will finish
this chapter compiling experimental results on heavy and light neutrino couplings.
• Chapter 6: At last, we will finish the thesis presenting our conclusions and future
prospects.
I hope this thesis is of interest to the reader, then all the enthusiasm and effort put
into it will be rewarded.
10
Preface
Prefaci
La fı́sica de neutrins es troba avuı́ en dia en una posició privilegiada dins d’un camp tan
apassionant com és el de la fı́sca de partı́cules. Des del descobriment de les oscil·lacions
de neutrins per part de Super-Kamiokande [1] a l’any 1998, la porta de la fı́sica més
enllà del Model Estàndar (SM a partir d’ara) ha sigut oberta. Aquest fet implica que els
neutrins han de ser massius [2] en contraposició al que suposa el Model Estàndard. No
obstant, açò no és una sorpresa completament, sinó que ja venia sent apuntada a partir
d’observacions teòriques i experimentals en les dues dècades prèvies al descobriment del
fenomen oscil·latori:
• En principi, no és necessari introduir la massa del neutrı́ al Model Estàndard. Però
aquesta no és una caracterı́stica que sorgeixi degut a una simetria (com és el cas del
fotó), sinó que s’elegeix per simplicitat, evitant, per exemple, introduir el neutrı́dret.
• La gran majoria de models d’unificació requerixen la presència de la massa del
neutrı́.
• El dèficit observat experimentalment en els fluxos de neutrins solars i atmosfèrics
podria ser explicat pel fenomen d’oscil·lacions de neutrins, mecanisme que requereix
l’existència de la massa dels neutrins i la mescla entre ells.
Com a conseqüència d’aquest nou camı́ obert pels neutrins, es podrien observar processos que es troben prohibits al Model Estàndard [5]:
• Les oscil·lacions de neutrins impliquen la mescla dels tres estats d’aquests; per tant
els nombres leptònics generacionals no seran vàlids com a simetries globals i es
podrien observar desintegracions on es viole el nombre de sabor leptònic (LFV):
– µ → eγ ,
– τ → eγ ,
11
12
Preface
– µ → 3e .
• Si els neutrins són partı́cules de Majorana, la desintegració doble beta sense neutrins
(0νββ) i altres processos on es viole el nombre leptònic podrien observar-se. Fins
que tal observació històrica es produı̈sca, el debat sobre la natura dels neutrins com
a partı́cules de Dirac o Majorana, continuarà obert.
• La cosmologia és sensible a l’existència de neutrins massius, els quals poden afectar
al fons còsmic de microones i a altres observables cosmològics.
• El neutrı́, com a partı́cula massiva, podria desintegrar-se:
– να → να 0 γ ,
– να → να0 να00 να000 , on tres neutrins són produı̈ts a l’estat final. Aquests podrien
ser del mateix tipus o no,
– να → να0 J , on J és un majoron 2 .
Com ja hem mencionat, cap dels fenòmens descrits en els paràgrafs anteriors es troba
contemplat al Model Estàndard. Per tant, models més enllà del Model Estàndard són
necessaris per a introduir la massa del neutrins, podent implicar l’existència de noves
interaccions amb la matèria: les Interaccions No Estàndard (NSI a partir d’ara). En
aquesta tesi estudiarem algunes d’aquestes noves interaccions des d’un punt de vista
fenomenològic. Per aconseguir-ho, utilitzarem les dades de diferent experiments de neutrins, obtenint aixı́ una major informació sobre les NSI a partir d’un anàlisi combinat
d’aquestes. Per a aquest propòsit, la tesi s’organitza de la següent forma:
• Capı́tol 1: Iniciarem el nostre treball amb una revisió de les caracterı́stiques més
importants dels neutrins dins del Model Estàndard. Junt amb aquestes, es presentaran també altres propietats més enllà del SM, com són la massa i les oscil·lacions
de neutrins.
• Capı́tol 2: En aquest capı́tol realitzarem una introducció a les NSI, explicant la
seva aparició a la teoria i les seves principals caracterı́stiques. Junt amb aquesta
explicació genèrica, discutirem els models on apareixen de forma espontània i la seva
influència sobre les probabilitats d’oscil·lació de neutrins i els experiments.
2
El majoron J és un bosó de Goldstone que apareix en models on la ruptura del nombre leptònic és
espontània [4].
13
• Capı́tol 3: El principal objectiu d’aquest capı́tol és mostrar que, encara que jugant
un paper secundari, les NSI encara tenen influència en la solució del problema dels
neutrins solars. Combinarem les dades d’experiments solars, reactor i accelerador,
per aixı́ obtindre lı́mits dels paràmetres que determinen la força d’aquestes NSI.
• Capı́tol 4: En aquesta part de la tesi obtindrem cotes als paràmetres de NSI implicats en interaccions del neutrı́ muònic amb quarks. Per a aquest comès utilitzarem
els resultats de la reanàlisis de NuTeV realitzats per dues col·laboracions en cojunt
amb dades d’experiments d’accelerador i d’atmosfèrics.
• Capı́tol 5: La no unitarietat de la matriu de mescla dels neutrins lleugers és
l’exemple més simple de NSI. En aquest capı́tol descriurem un formalisme que ens
permetrà treballar de forma senzilla amb models que consideren més de tres neutrins
(com els models seesaw), separant nova fı́sica d’aquella que és estàndard. Acabarem
aquest capı́tol fent una recopilació dels resultats experimentals de l’acoblament d’un
neutrı́ pesat amb un altre lleuger.
• Capı́tol 6: Finalment, per acabar la tesi presentarem les nostres conclusions i
perspectives a futur.
Espere que aquesta tesi siga de l’interés del lector, d’aquesta forma tot el treball,
il·lusió i esforç posats en ella seran recompensats.
14
Preface
List of publications
This thesis is based on the following publications carried out with the supervisors of the
thesis and several collaborators:
• F. J. Escrihuela, O. G. Miranda, M. A. Tórtola and J. W. F. Valle,
“Constraining nonstandard neutrino-quark interactions with solar, reactor and accelerator data”.
Phys. Rev. D 80, 105009 (2009), arXiv:0907.2630 [hep-ph].
• F. J. Escrihuela, O. G. Miranda, M. Tórtola and J. W. F. Valle,
“Global constraints on muon-neutrino non-standard interactions”.
Phys. Rev. D 83, 093002 (2011), arXiv:1103.1366 [hep-ph].
• F. J. Escrihuela, D. V. Forero, O. G. Miranda, M. Tórtola and J. W. F. Valle,
“On the description of non-unitary neutrino mixing”.
Phys. Rev. D 92, 053009 (2015), arXiv:1503.08879 [hep-ph].
The results of these works were presented in several conferences and they have been
published as proceedings:
• F. J. Escrihuela, O. G. Miranda, M. A. Tórtola and J. W. F. Valle,
“Constraining nonstandard neutrino-quark interactions with solar, reactor and accelerator data”.
J. Phys. Conf. Ser. 259, 012091 (2010).
Talk presented at PASCOS 2010, Valencia (Spain).
• F. J. Escrihuela,
“Status of neutrino-quark NSI parameters”.
15
16
List of publications
J. Phys. Conf. Ser. 375, 042046 (2012).
Talk presented at TAUP 2011, Munich (Germany).
• F. J. Escrihuela,
“Analysis of non-standard neutrino-quark interactions”.
AIP Conf. Proc. 1420, 97 (2012).
Talk presented at EAV 2011, CINVESTAV, Mexico City (Mexico).
• F. J. Escrihuela,
“Constraining right-handed neutrinos”.
Conference: C14-07-02, arXiv:1505.01097 [hep-ph].
Talk presented at ICHEP 2014, Valencia (Spain).
Contents
Preface
7
List of publications
14
1 Neutrinos in the Standard Model and beyond
21
1.1
1.2
1.3
1.4
1.5
1.6
Brief description of the Standard Model
. . . . . . . . . . . . . . . . . . . 23
1.1.1
Electroweak theory . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.1.2
A touch of QCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Neutrinos in the Standard Model . . . . . . . . . . . . . . . . . . . . . . . 28
1.2.1
Origin and number of neutrinos . . . . . . . . . . . . . . . . . . . . 29
1.2.2
Chirality and helicity of neutrinos . . . . . . . . . . . . . . . . . . . 31
1.2.3
Massless neutrinos . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Massive neutrinos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
1.3.1
Dirac vs Majorana particles . . . . . . . . . . . . . . . . . . . . . . 34
1.3.2
Producing the neutrino mass . . . . . . . . . . . . . . . . . . . . . . 37
Neutrino mass models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.4.1
Type I+II seesaw . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
1.4.2
Inverse seesaw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
1.4.3
Linear seesaw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
1.4.4
Left-right symmetric model . . . . . . . . . . . . . . . . . . . . . . 42
1.4.5
Radiative models . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
1.4.6
Supersymmetry as origin of neutrino mass . . . . . . . . . . . . . . 43
Neutrino oscillations in vacuum . . . . . . . . . . . . . . . . . . . . . . . . 44
1.5.1
The neutrino oscillation probability in vacuum . . . . . . . . . . . . 45
1.5.2
Oscillations with two neutrinos . . . . . . . . . . . . . . . . . . . . 49
1.5.3
Oscillations with three neutrinos
. . . . . . . . . . . . . . . . . . . 50
Neutrino oscillations in matter . . . . . . . . . . . . . . . . . . . . . . . . . 52
17
18
Contents
1.7
1.6.1
Neutrino evolution in matter . . . . . . . . . . . . . . . . . . . . . . 52
1.6.2
Oscillations in a medium of constant density . . . . . . . . . . . . . 55
1.6.3
Oscillations in an adiabatic medium . . . . . . . . . . . . . . . . . . 56
1.6.4
Oscillations in a non-adiabatic medium . . . . . . . . . . . . . . . . 58
Oscillation data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2 Non-Standard Interactions
2.1
2.2
2.3
2.4
63
Parameterization of NSI . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
NSI from extended models
. . . . . . . . . . . . . . . . . . . . . . . . . . 65
2.2.1
Models with extra neutral gauge bosons . . . . . . . . . . . . . . . 65
2.2.2
Seesaw models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.2.3
SUSY with R-parity violation . . . . . . . . . . . . . . . . . . . . . 69
2.2.4
Models with leptoquarks . . . . . . . . . . . . . . . . . . . . . . . . 70
NSI phenomenology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
2.3.1
NSI in the source and detector . . . . . . . . . . . . . . . . . . . . . 71
2.3.2
Neutrino propagation with NSI . . . . . . . . . . . . . . . . . . . . 72
NSI and neutrino experiments . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.4.1
NSI in atmospheric neutrino experiments . . . . . . . . . . . . . . . 77
2.4.2
NSI in accelerator neutrino experiments . . . . . . . . . . . . . . . 78
2.4.3
NSI in solar neutrino experiments . . . . . . . . . . . . . . . . . . . 78
2.4.4
NSI in short baseline experiments . . . . . . . . . . . . . . . . . . . 81
3 Robustness of solar neutrino oscillations
3.1
3.2
3.3
83
Solar and KamLAND restrictions on NSI . . . . . . . . . . . . . . . . . . . 85
3.1.1
The solar and KamLAND data . . . . . . . . . . . . . . . . . . . . 85
3.1.2
Effects of NSI in neutrino propagation . . . . . . . . . . . . . . . . 86
3.1.3
Constraints on vectorial NSI . . . . . . . . . . . . . . . . . . . . . . 91
3.1.4
NSI effects in neutrino detection . . . . . . . . . . . . . . . . . . . . 92
Analysis including CHARM data . . . . . . . . . . . . . . . . . . . . . . . 95
3.2.1
Constraints on non-universal NSI from CHARM . . . . . . . . . . . 95
3.2.2
Constraints on NSI from a combined analysis . . . . . . . . . . . . 97
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4 Probing νµ NSI with accelerator and atmospheric data
101
4.1
Neutrino-nucleon scattering and measurements of sin2 θW . . . . . . . . . . 102
4.2
The NuTeV data and constraints on NSI . . . . . . . . . . . . . . . . . . . 104
Contents
4.3
4.4
4.5
19
NSI in the CHARM and CDHS experiments . . . . . . . . . . . . . . . . . 108
Combining with atmospheric neutrino data . . . . . . . . . . . . . . . . . . 110
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
5 Non-unitary lepton mixing matrix
5.1 A new way to describe extra heavy leptons
5.2 NHL and non-unitary neutrino mixing . .
5.3 Constraints from universality tests . . . .
5.4 Neutrino oscillations with non-unitarity . .
5.5 Neutrino oscillations and universality . . .
5.6 Current constraints on NHL . . . . . . . .
5.6.1 Neutrinoless double beta decay . .
5.6.2 Charged lepton flavor violation . .
5.7 Summary . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
115
. 117
. 119
. 120
. 123
. 130
. 132
. 135
. 137
. 138
6 Conclusions
139
A NSI in a two-neutrino scheme
147
B Neutrino mixing and heavy isosinglets
149
C Relevant neutrino experiments
C.1 Reactor experiments . . . . . . .
C.2 Solar experiments . . . . . . . . .
C.3 Accelerator experiments . . . . .
C.4 Experiments giving constraints on
Bibliography
. . .
. . .
. . .
NHL
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
157
. 157
. 161
. 164
. 166
171
20
Contents
Chapter 1
Neutrinos in the Standard Model
and beyond
The Standard Model of particles is the most successful theory describing the interactions
of elementary particles. This quantum field theory describes the strong, weak and electromagnetic interactions of elementary particles using the following components summarized
in Table 1.1:
• Fermions, which are particles with spin 12 . They are classified into three families:
– First family: Up and down quarks, electron and electron-neutrino. These are
the components of the ordinary matter.
– Second family: Charm and strange quarks, muon and muon-neutrino.
– Third family: Top and bottom quarks, tau and tau-neutrino.
• Gauge bosons, correspond to particles with spin 1, responsible for carrying the
three interactions mentioned above:
– Photon: It is the gauge boson of the electromagnetic interaction. It is a
massless particle.
– W + , W − and Z 0 : These particles mediate the weak interaction, W + and
W − for charged-current weak interactions and Z 0 for neutral-current weak
interactions. Unlike photons and gluons, these particles have a mass.
– Gluon: This gauge boson mediates strong interactions between quarks. As
the photon, it is a massless particle.
21
22
Chapter 1. Neutrinos in the Standard Model and beyond
Table 1.1: Standard Model components.
Fermions
Quarks
Leptons
Boson
First family
Second family
Third family
u (up)
c (charm)
t (top)
d (down)
s (strange)
b (bottom)
νe (electron-neutrino) νµ (muon-neutrino) ντ (tau-neutrino)
e (electron)
µ (muon)
τ (tau)
Electromagnetism
γ (photon)
Weak
W , W − and Z 0
+
Strong
AC (gluon)
Higgs boson
• Higgs boson: This particle is the last ingredient in the Standard Model of particles.
It is responsible for the mass of particles and interacts with them through the socalled Higgs mechanism. It is a particle with spin 0, and a mass of mH ≈ 125 GeV .
This model is supported by a large number of experimental results, as for example [6]:
• Particle predictions: W , Z, gluons, top quark, charm quark...
• Couplings constants to Z 0 gauge boson in e− e+ → f¯f processes.
• Branching ratio in Z 0 → f¯f , W − → ν̄l l and W − → ūi dj .
• Leptonic universality.
• Z invisible decay width Z 0 → να ν̄α .
• Higgs boson.
• ...
However, despite this great success, there are several open questions which have not been
answered yet:
• The mechanism which generates neutrino masses.
• Baryonic asymmetry of the Universe.
1.1. Brief description of the Standard Model
23
• Dark matter nature.
• The violation of CP symmetry in the leptonic sector.
• Neutrino nature, Dirac or Majorana?.
• The neutrino mass hierarchy problem.
• Proton stability.
• ....
Finding solutions to these problems has led to the proposal of “new physics” beyond the
Standard Model. Neutrinos are an important motivation for physics beyond the Standard
Model, as we will see in this thesis.
1.1
Brief description of the Standard Model
The Standard Model is a gauge theory mathematically described by the local symmetry
group SU (3)C × SU (2)L × U (1)Y :
• SU (3)C stands for the description of strong interaction through the color quantum
number. This symmetry group has eight generators which correspond to the eight
gluons that mediate the strong force.
• SU (2)L represents the weak interaction whose local symmetry is the weak isospin.
Only left-handed particles with weak isospin I = 12 couple to weak bosons. As this
symmetry group has three generators, the number of gauge bosons is three too, W + ,
W − and Z 0 as we have seen above.
• U (1)Y is the symmetry group for the electromagnetic interaction. In this case, the
local symmetry is the hypercharge (Y ), having only one generator which is related
with the photon, the unique gauge boson related to this group.
With these ingredients, Sheldon Glashow, Abdus Salam and Steven Weinberg were
able to unify the weak and electromagnetic interaction in the so-called electroweak interaction. This interaction is described by the SU (2)L × U (1)Y symmetry group, having four
gauge bosons, three corresponding to the three generators of SU (2)L , W i , and one gauge
boson coming from the electromagnetic U (1)Y group, B. After spontaneous symmetry
24
Chapter 1. Neutrinos in the Standard Model and beyond
breaking, the usual gauge bosons, W + , W − , Z 0 and γ, which mediate the electroweak
interaction, arise as a combination of the first ones.
On the other hand, strong interaction cannot be combined with other force, that is
why SU (3)C is studied separately of SU (2)L × U (1)Y .
Now, we will take a quick look of the mathematical description of each group, focusing
mainly on the description of the electroweak interaction, related, among others, with
neutrinos.
1.1.1
Electroweak theory
The electroweak model is described by the SU (2)L × U (1)Y group. SU (2)L is the group
of weak isospin (I), which is a quantum number assigned only to left-handed quarks and
leptons of each generation. This violates parity symmetry giving rise to a V-A theory.
Keeping this in mind, the leptons can be grouped into doublets of chiral left-handed fields
and singlets of right-handed fields, as we show in Table 1.2.
As we have said before, the SU (2)L has three generators denoted by
1
Ik = σk ,
2
(1.1)
where k goes from 1 to 3 and σk are the isospin version of the Pauli matrices:
σ1 =
0 1
1 0
!
,
σ2 =
0 −i
i 0
!
,
σ3 =
1 0
0 −1
!
.
(1.2)
These generators satisfy the angular momentum commutation relations, as spin does:
[Ik , Ij ] = ikjl Il .
(1.3)
Besides SU (2)L , we find the symmetry group U (1)Y which is known as the weak
hypercharge group and its generator is the weak hypercharge Y . This operator relates
the weak isospin I3 with the charge operator Q through the Gell-Mann-Nishijima relation:
Q = I3 +
Y
.
2
(1.4)
This expression illustrates the unification of weak and electromagnetic interactions.
Once we have the basic ingredients of electroweak theory, we are going to study the
lagrangian of the model. For this purpose and for simplicity, we will take into account
1.1. Brief description of the Standard Model
25
Table 1.2: Quantum numbers of leptons and quarks.
Lepton Doublet LL =
Lepton Singlet
Quark Singlet
QL =
uR
dR
uL
dL
I3
Y
Q
1/2
1/2
−1/2
−1
0
−1
0
0
−2
−1
1/2
1/2
−1/2
1/3
2/3
−1/3
0
0
4/3
−2/3
2/3
−1/3
eR
Quark Doublet
νeL
eL
I
only the first generation leptons (eL , νeL and eR ) considering them as massless particles 1 ;
we begin our discussion assuming that leptons have no couplings with electroweak gauge
bosons. With these conditions, the lagrangian for these free Dirac fields can be written
as
!
ν
eL
L = (ν̄eL , ēL )(iγ µ ∂µ )
+ ēR iγ µ ∂µ eR .
(1.5)
eL
Note that we have respected the doublet-singlet notation. This lagrangian is not invariant
under weak isospin transformations 2 , but we can introduce gauge vector fields to compensate it. As we said before, we find three vectors fields in SU (2)L as generators of the
symmetry group, W i µ (with i = 1, 2, 3). Including these fields, the lagrangian becomes
1
L = T r(Wµρ W µρ ) + (ν̄eL , ēL )iγ µ (∂µ + igWµ )
2
νeL
eL
!
+ ēR iγ µ ∂µ eR .
(1.6)
The gauge bosons W Q have electromagnetic charge and they are defined as a linear
1
We know from neutrino oscillations that the neutrino is a massive particle in the SU (2)L × U (1)Y
description of the electroweak theory. But, in order to simplify the discussion, here the neutrino is chosen
without mass and only left-handed.
2
Weak isospin transformations can be considered as space rotations of a vector.
26
Chapter 1. Neutrinos in the Standard Model and beyond
combination of W i µ as:
1
Wµ+ = √ (Wµ1 − iWµ2 ) ,
2
1
Wµ− = √ (Wµ1 + iWµ2 ) ,
2
Wµ0 = Wµ3 .
(1.7)
Introducing these fields in Eq. (1.6), we obtain the following lagrangian for the leptonboson coupling term,
σ
L = −g(ν̄eL , ēL )γ µ Wµ
2
νeL
eL
!
(1.8)
!
!
√
0
+
W
2
W
ν
1
eL
µ
µ
√
= −g(ν̄eL , ēL )γ µ
2
2 Wµ− −Wµ0
eL
h
i
√
√
g
= − Wµ0 (ν̄eL γ µ νeL − ēL γ µ eL ) + 2 Wµ+ ν̄eL γ µ eL + 2 Wµ− ēL γ µ νeL ,
2
where σ are the Pauli matrices in Eq. (1.2) and the weak isospin components are defined as
Ik = 21 σk . We can rewrite Eq. (1.8) in a more usual form using the left-chirality projector
γ 5 3,
1
ψ̄L γ µ ψL = ψ̄γ µ (1 − γ 5 )ψ ,
(1.9)
2
giving rise to
√
g
L = − Wµ0 (ν̄e γ µ (1 − γ 5 )νe − ēγ µ (1 − γ 5 )e) + 2 Wµ+ ν̄e γ µ (1 − γ 5 )e
4
√
−
µ
5
+
2 Wµ ēγ (1 − γ )νe .
(1.10)
As we can see from the equations above, only left-handed leptons interact with weak
gauge bosons, as expected.
Since we are talking about electroweak interaction, one extra-component must be
introduced in the lagrangian, electromagnetism. We could think that Wµ0 is associated
with the photon, but this is not possible because Wµ0 and the photon have different
couplings. Therefore, the electromagnetic interaction is inserted in the lagrangian through
the hypercharge group U (1)Y , where, as we have said in Eq. (1.4), charge and isospin are
related. In this group, the vector field is Bµ with the corresponding coupling constant g 0 .
3
It is defined in Sec. 1.2.2.
1.1. Brief description of the Standard Model
27
Both neutral vector fields combine as
Zµ = p
Aµ = p
1
g 2 + g 02
1
g 2 + g 02
gWµ0 − g 0 Bµ ,
(1.11)
g 0 Wµ0 + gBµ ,
(1.12)
leading to gauge bosons of weak neutral-current and electromagnetism respectively. Defining the weak mixing angle θW in terms of the coupling constants g and g 0 ,
g0
sin θW = p
,
g 2 + g 02
cos θW = p
g
g2
+ g 02
,
(1.13)
the expressions above, Eqs. (1.11) and (1.12), are simplified as
Zµ = cos θW Wµ0 − sin θW Bµ ,
(1.14)
Aµ = sin θW Wµ0 + cos θW Bµ .
(1.15)
Replacing Zµ and Aµ in Eq. (1.8), it results in
g
L = − √ Wµ+ ν̄eL γ µ eL + Wµ− ēL γ µ νeL
2
p
1
1
µ
µ
2
µ
µ
2
02
g + g Zµ ν̄eL γ νeL − ēL γ eL + sin θW (ēL γ eL + ēR γ eR )
−
2
2
0
gg
+ p
Aµ (ēL γ µ eL + ēR γ µ eR ) ,
(1.16)
2
02
g +g
where the vector field Aµ is only coupled to charged leptons, so it can be associated with
the photon, whereas Wµ± and Zµ represent the gauge bosons of weak charged-current and
weak neutral-current respectively.
In order to relate the electromagnetism and weak interactions, along with the Eq. (1.4),
we have
gg 0
p
= e,
(1.17)
g 2 + g 02
which leads to the relation between the weak mixing angle and the electromagnetic charge:
sin θW =
e
.
g
(1.18)
Finally, we may write the complete lagrangian including electromagnetism, charged-
28
Chapter 1. Neutrinos in the Standard Model and beyond
current and neutral-current weak interactions, respectively:
1
1
µ
µ
+
µ
−
µ
L = −e Aµ Jem + √
Wµ ν̄eL γ eL + Wµ ēL γ νeL +
Zµ JN C ,
sin θW cos θW
2 sin θW
(1.19)
where
µ
Jem
= − (ēL γ µ eL + ēR γ µ eR ) = −ēγ µ e ,
1
1
µ
ν̄eL γ µ νeL − ēL γ µ eL − sin2 θW Jem
.
JNµ C =
2
2
1.1.2
(1.20)
(1.21)
A touch of QCD
For the sake of completeness, we just mention QCD which is described by the symmetric
group SU (3)C , as already stated. Quantum Chromodynamics is the gauge theory that
describes the strong interaction between colored quarks and is mediated by eight gluons
carrying color. The QCD lagrangian is [7]
L=
X
q
1 A A µν
C
ψ̄q,a (iγ µ ∂µ δab − gs γ µ tC
,
ab Aµ − mq δab )ψq,b − Fµν F
4
(1.22)
where
• γ µ are the Dirac matrices,
• ψq,a stand for quark-field spinors of mass mq , with a = 1, 2, 3 which is the number
of colors,
• AC
µ correspond to the gluon fields, where C goes from 1 to 8, the number of gluons
in QCD,
• tcab are the eight generators of SU (3)c ,
• gs is the QCD coupling constant,
A
A
B C
• Fµν
= ∂µ AA
ν −∂ν Aµ −gs fABC Aµ Aν , with fABC as the structure constants of SU (3)C .
1.2
Neutrinos in the Standard Model
Neutrinos are electrically neutral particles of spin 12 . There are three flavors of neutrinos
in the SM, νe , νµ and ντ (Table 1.1), which are left-handed, being associated with a lepton
1.2. Neutrinos in the Standard Model
29
of each family as we saw in Sec. 1.1.1 (Table 1.2). We also find their three antiparticles,
the antineutrinos ν̄e , ν̄µ and ν̄τ , which are right-handed particles. Right-handed neutrinos
and left-handed antineutrinos have not been experimentally found, which corresponds to
particles without mass in the Standard Model 4 . In the rest of the chapter, we are going
to deepen slightly in neutrino properties in order to provide the reader a first insight into
neutrino theory.
1.2.1
Origin and number of neutrinos
Neutrinos and antineutrinos are particles produced in several processes as, for instance:
• Nuclear β ± decay,
A(Z, N ) → A(Z + 1, N − 1) + e− + ν̄e ,
A(Z, N ) → A(Z − 1, N + 1) + e+ + νe .
(1.23)
• Muon decays,
µ+ → e+ + ν̄µ + νe ,
µ− → e− + νµ + ν̄e .
(1.24)
• Pion decays,
π + → e + + νe ,
π − → e− + ν̄e ,
π + → µ+ + νµ ,
π − → µ− + ν̄µ .
(1.25)
• Tau decays,
τ − → µ− + ν̄µ + ντ ,
τ + → µ+ + νµ + ν̄τ .
4
(1.26)
Particles require both chiralities to produce mass in the Standard Model, as we will see in Sec. 1.2.3.
30
Chapter 1. Neutrinos in the Standard Model and beyond
Figure 1.1: Neutrino weak interactions with leptons. The left diagram stands for chargedcurrent weak interactions through W ± , whereas the right diagram represents neutralcurrent weak interactions mediated by Z 0 .
• Solar neutrino production,
p+ + p+ →
2
H + e + + νe ,
p + + e − + p+ →
2
H + νe ,
He + p+ →
4
He + e+ + νe ,
Be + e− →
2
Li + νe ,
8
8
Be + e+ + νe .
3
7
B →
(1.27)
Neutrinos interact with charged leptons through charged-current weak interactions, Fig. 1.1,
coupling with bosons W ± in reactions involving
W ± → lα± + να (ν̄α ) ,
(1.28)
where α indicates the flavor of the lepton and the neutrino. Similarly, they also participate
in processes mediated by the Z 0 boson, the so-called neutral-current weak interactions
Fig. 1.1. These are elastic or quasi-elastic scattering processes and Z 0 decays,
Z 0 → να ν̄α .
Notice that these couplings were given before in Eqs. (1.16) and (1.19).
(1.29)
1.2. Neutrinos in the Standard Model
31
Following with the Z 0 decay, it allowed to determine the number of light neutrinos.
Indeed, neutrinos from Z 0 are not detected directly, but the presence of neutrinos is known
from the difference between the total decay width of the Z 0 boson and widths of visible
particles after decaying,
Γinv = Γtot − Γvis = 499.0 ± 1.5 MeV .
(1.30)
This invisible width should correspond to the decay into a pair ν ν̄ of any flavor. So, considering that the width of Z 0 decaying to one pair is Γν ν̄ = 167.2 MeV, we can translate this
into the number of light active neutrinos species (with their corresponding antineutrinos)
as [6, 8]
Γinv
Nν =
= 2.984 ± 0.082 ,
(1.31)
Γν ν̄
almost three light neutrino flavors, as it is assumed in the Standard Model.
1.2.2
Chirality and helicity of neutrinos
In Quantum Field Theory, fermions can be described by four-component spinors ψ which
obey the Dirac equation,
µ ∂
iγ
− m ψ = 0,
(1.32)
∂xµ
where γ µ are the Dirac matrices
γ0 =
5
0 1
1 0
!
,
γi =
0 σi
−σi 0
!
(1.33)
and σi are the Pauli matrices, given already in Eq. (1.2). We want to stress the importance
of the combination of Dirac matrices in
!
−1 0
γ5 = iγ0 γ1 γ2 γ3 =
,
(1.34)
0 1
because, as we will see in the following lines, this matrix will give rise to the chirality
operator.
Using the Eq. (1.32) and the Dirac matrices, taking into account that γi = γ0 γ5 σi ;
we obtain the Dirac equation in a convenient form in terms of γ5 , decoupled in two
5
Other conventions of these matrices are also used in the literature.
32
Chapter 1. Neutrinos in the Standard Model and beyond
expressions:
∂
∂
i 0 (1 + γ5 ) + iσi
(1 + γ5 ) − mγ0 (1 − γ5 ) ψ = 0 ,
∂x
∂xi
∂
∂
i 0 (1 − γ5 ) − iσi
(1 − γ5 ) − mγ0 (1 + γ5 ) ψ = 0 .
∂x
∂xi
(1.35)
(1.36)
From the equations above, we can define two projection operators given by
1
PL = (1 − γ5 ) ,
2
1
PR = (1 + γ5 ) ,
2
(1.37)
getting the left-handed and right-handed component of the fermion ψ as
ψL = PL ψ,
ψR = PR ψ
and
PL ψR = PR ψL = 0.
(1.38)
As we mentioned before, γ5 plays a key role as a chirality operator
γ5 ψL,R = ∓ψL,R ,
(1.39)
where ∓1 correspond to the chirality eigenvalues and ψL,R are the chiral projections of ψ.
Any spinor ψ can be written in terms of its projections as follows
ψ = (PL + PR )ψ = PL ψ + PR ψ = ψL + ψR .
(1.40)
Chirality is related with helicity, in such a way that, if a fermion ψ is a massless particle,
both properties are the same. Now we will see it.
Combining Eqs. (1.35), (1.36) and (1.40), we may rewrite the Dirac equation in terms
of left and right-handed components, obtaining:
∂
∂
ψR = mγ0 ψL ,
i 0 − iσi
∂x
∂xi
∂
∂
i 0 + iσi
ψL = mγ0 ψR .
∂x
∂xi
(1.41)
(1.42)
In the case of vanishing mass, the equations can be rewritten as
∂
∂
ψR = iσi
ψR ,
0
∂x
∂xi
∂
∂
i 0 ψL = −iσi
ψL ,
∂x
∂xi
i
(1.43)
(1.44)
1.2. Neutrinos in the Standard Model
33
Figure 1.2: Helicity for neutrino and antineutrino, which is coincident with chirality in
the Standard Model, where mν = 0.
which are “equal” to the Schrödinger equation, with x0 = t and ~ = 1. Using the
∂
and pi = −i ∂x∂ i , Eqs. (1.41) and (1.42) acquire a more compact form
definitions E = i ∂t
EψL,R = ±σi pi ψL,R .
(1.45)
This equation defines a new operator, the helicity ĥ, as
ĥ =
σ·p
,
|p|
(1.46)
which is the projection of the spin over the linear momentum, as we may see in Fig. 1.2.
Notice from Eq. (1.45) that ψL represents a spinor with ĥ = +1 for particles and ĥ = −1
for antiparticles. On the other hand, ψR stands for a spinor with helicity ĥ = −1 for
particles and ĥ = +1 for antiparticles. Therefore, if we neglect the fermion mass, chirality
and helicity are the same.
So that, we may affirm that, in the Standard Model, where the neutrino is a massless
particle, the interacting neutrino is always left-handed, and therefore ĥ = +1,
1
(1 − γ5 )ν = νL ,
2
(1.47)
while the antineutrino is always right-handed, with ĥ = −1.
1.2.3
Massless neutrinos
As we will briefly discuss, in the Standard Model neutrinos are particles without mass
and this fact is related with their chirality, studied in the previous section. From the
Higgs mechanism, fermions acquire mass through their couplings to the Higgs field. The
34
Chapter 1. Neutrinos in the Standard Model and beyond
resulting coupling is known as the Yukawa coupling. Here, we take the lagrangian for the
first family of leptons for illustration:
"
νeL
eL
L = −λe (ν̄eL , ēL ) φ0 eR + ēR φ†0
1
1
= −λe ēL √ veR + ēR √ veL
2
2
1
= −λe √ v (ēL eR + ēR eL )
2
1
= −λe √ v (ēe) .
2
!#
(1.48)
As we can appreciate, only particles with both (left and right) chiralities couple to the
Higgs boson and thus obtain their mass. In contrast, neutrinos remain massless because
they do not couple to the Higgs field (no evidence of right-handed neutrino or left-handed
antineutrino has been found). But neutrino mass is a reality, so mechanisms beyond the
Standard Model are introduced to generate the neutrino mass, as we will see in the next
section.
1.3
Massive neutrinos
In the Standard Model, neutrinos are taken as particles without mass. However, after the
discovery of neutrino oscillations [1] (Sec. 1.5), the situation has changed and introducing
neutrino masses is necessary. Nevertheless, the absolute value of the neutrino mass has
not been measured yet, because only the mass differences between neutrinos and the
mixing angles enter in neutrino oscillations.
Along with this, there are other mysteries that accompany neutrino masses, as the
mechanism responsible for them and their small size compared with the rest of fermions.
Answering these and other related questions has led people to develop many models in
order to shed light on these mysteries. In what follows, we will discuss the neutrino mass,
taking into account the particle nature (Dirac or Majorana), as well as the mechanism to
originate it.
1.3.1
Dirac vs Majorana particles
Let us start considering the Dirac lagrangian for free particles:
L = ψ̄ (iγ µ ∂µ − m) ψ ,
(1.49)
1.3. Massive neutrinos
35
where the four-component Dirac spinor can be written in terms of two-component lefthanded spinors χ and ξ as [3, 9],
ψ=
χ
iσ2 ξ ∗
!
=
χ
ξ¯
!
,
(1.50)
with σ2 as one of the Pauli matrices defined in Eq. (1.2). Therefore, we may expand
Eq. (1.49) in terms of two-component spinor (Eq. (1.50)) obtaining the following expression:
L = iξσ µ ∂µ ξ¯ + iχ̄σ̄ µ ∂µ χ − m ξχ + χ̄ξ¯
¯ µ ∂µ ξ + iχ̄σ̄ µ ∂µ χ − m ξχ + χ̄ξ¯ ,
= iξσ̄
(1.51)
where the following definitions of the Pauli matrices have been used [3],
σ µ ≡ (1, ~σ ) ,
σ̄ µ ≡ (1, −~σ ) .
(1.52)
Notice that in Eq. (1.51) the first two terms represent the kinetic energy, while the last
one stands for the Dirac mass term given by
LmD = −m ξ T iσ2 χ + χ† iσ2 ξ ∗ ,
(1.53)
where we have used the condition ψ̄ = ψ † γ 0 .
If we identify the two-component spinors χ and ξ with the chiral-components of some
field ψ [3, 9],
!
!
ψL
χ
ψ=
=
,
(1.54)
ψR
ξ¯
we see that both chiralities (left and right) are necessary in order to produce a Dirac mass
term, as it can be observed from Eq. (1.53). But as we already mentioned in Sec. 1.2.2,
the SM only includes left-handed neutrinos νL ≡ χ, avoiding the right-handed component,
and right-handed antineutrinos νRc , avoiding the left-handed component in this case. As
a consequence, neutrinos remain massless in the SM, as we have discussed in Sec. 1.2.3.
Note that one can rewrite the two-component spinors in Eq. (1.50) as
1
χ = √ (ρ1 + iρ2 ) ,
2
1
ξ = √ (ρ1 − iρ2 ) ,
2
(1.55)
36
Chapter 1. Neutrinos in the Standard Model and beyond
so Eq. (1.51) becomes
2
X
2
1 X
L=i
ρ̄a σ̄ ∂µ ρa − m
(ρa ρa + ρ̄a ρ̄a ) .
2 a=1
a=1
µ
(1.56)
Each of these ρ fields by itself represents a Majorana particle:
ρ = ρc = C ρ̄T ,
(1.57)
where C is defined as the charge conjugation operator, which connects a particle p(x, t)
with its antiparticle p̄(x, t) as [10–12]:
C|p(x, t)i = ηc |p̄(x, t)i ,
(1.58)
with |ηc | = 1. It can be defined explicitly as
C=
−iσ2 0
0
iσ2
!
.
(1.59)
Hence, we can separate Eq. (1.56) in two pieces:
1
L = iρ̄ σ̄ µ ∂µ ρ − m (ρρ + ρ̄ρ̄) ,
2
(1.60)
getting the following Majorana mass term,
1
LmM = − m ρT iσ2 ρ + h.c. .
2
(1.61)
From the whole discussion we can find that a four-component Dirac fermion is equivalent to two Majorana fermions of equal mass, represented by two-component spinors.
Hence, the description with a two-component spinor is the most basic representation of a
spin-1/2 fermion and we consider it more general in order to discuss the neutrino mass.
Using a four-component spinor description, we define a Majorana fermion as [3, 9]
c
T
ψM = ψM
= C ψ̄M
.
(1.62)
We can see from Eq. (1.62) that both two-component spinors are the same for a Majorana
1.3. Massive neutrinos
37
particle [3, 9]:
χ
iσ2 χ∗
ψM =
!
χ
χ̄
=
!
.
(1.63)
Notice that only neutral particles could be described as Majorana.
1.3.2
Producing the neutrino mass
Since neutrinos are massless particles in the Standard Model, extra components have to be
added in order to introduce the neutrino mass. If we want to produce massive neutrinos
without extending the lepton sector, it is necessary to introduce a SU (2)L Higgs triplet [9],
which is written in matrix form as
!
+
∆
++
√
∆
2
.
(1.64)
∆=
+
√
∆0 − ∆
2
This Higgs triplet will produce the following Yukawa terms in the weak lagrangian [9]:
−
1X
gab laT C −1 iσ2 ∆lb + h.c. ,
2 a,b
(1.65)
where C is the charge conjugation operator, l represents the SM doublet (Table 1.1) and
gab is the coupling constant. After spontaneous symmetry breaking, the Higgs triplet
acquires a vacuum expected value (vev in what follows), generating a Majorana mass
term as in Eq. (1.61). The following condition has to be verified by the Higgs triplet [13],
2
hφ0 i
h∆ i = −µ∆ 2 ,
M∆
0
(1.66)
arising from the minimization of the scalar potential. As neutrino masses are proportional
to gab h∆0 i, the smallness of neutrino mass will be conditioned by the Higgs triplet mass
M∆ .
Another way to generate neutrino masses is introducing extra components to the
fermion sector of the SM. For instance, SU (2)L isosinglets, represented by left-handed
spinor fields ρLa , can be added producing Majorana mass terms such as [9]
−
X
0 T
gab
ρLa C −1 ρLb + h.c.
(1.67)
a,b
In addition, a Dirac mass term can be induced through the coupling between the SM
38
Chapter 1. Neutrinos in the Standard Model and beyond
doublet l and the singlet ρL :
X
00 ¯
gab
la iσ2 φ∗ C ρ̄TLb + h.c. ,
(1.68)
a,b
with φ as the SM Higgs doublet.
In a very general model of neutrinos, we can find n neutrinos belonging to SU (2)L
doublets and described by two-component spinors ρn , along with m extra neutrinos as
SU (2)L singlets, ρm . The lagrangian of this model is a generalization of Eq. (1.60), which
becomes
n+m
X
1
L=
(1.69)
iρ̄α σ̄µ ∂ µ ρα − (Mαβ ρα ρβ + h.c.) ,
2
α,β
where α and β run from 1 to n + m. The mass matrix M is a symmetric
can be written in general as
!
ML MD
.
M=
MDT MR
6
matrix that
(1.70)
Here, the n × n ML submatrix comes from Eq. (1.65) where a Higgs triplet is present, the
m × m MR piece follows from Eq. (1.67) where extra SU (2)L isosinglets couple among
them, and the MD block stems from Eq. (1.68) as a consequence of the coupling between
SU (2)L doublets and singlets.
In the most general case, when all the submatrices described above are present, models
are called Type-I+II seesaw. But if the Higgs triplet is not present and ML = 0, they
will be named Type-I seesaw 7 . Finally, if a fermion triplet (TF ) is introduced as a mass
messenger, we have the Type-III seesaw [14]. For a more detailed analysis on this topic,
the reader should consult Refs. [3–5, 9, 12, 14] among others.
1.4
Neutrino mass models
The most popular mechanism to provide the neutrino mass is the seesaw mechanism [15].
Mainly it is based on the idea of generating the effective dimension-five operator O5 =
λLΦLΦ [16], Fig. 1.3 (where L represents a lepton doublet for each family and Φ is the
SM scalar doublet), by the interchange of heavy particles, that could be fermions (Type-I
and Type-III) or scalars (Type-II). This can be achieved in different ways, with several
multiplet contents and different gauge groups. The mass of the messenger determines the
6
7
The symmetry condition implies Mαβ = Mβα .
Notice that we are using the standard notation, but we can find different notation as in Ref. [9].
1.4. Neutrino mass models
39
Φ
Φ
L
L
Figure 1.3: Dimension five operator which leads to the neutrino mass [16].
scale where the global lepton number symmetry is violated due to the presence of new
physics:
hΦi2
,
(1.71)
m ν = λ0
MX
with λ0 as an unknown dimensionless constant.
The seesaw mechanism is an elegant way of, not only introducing neutrino masses, but
also explaining its smallness, since as the mass of the intermediate particles increases, the
neutrino mass decreases at fixed coupling. In this section we will review some significant
examples of these models.
1.4.1
Type I+II seesaw
In a (3 + 3) neutrino scheme 8 , with three left-handed and three right-handed neutrinos represented by νL and NLc respectively, the seesaw mass matrix can be built as in
Eq. (1.70) [9],
!
ML MD
M6 =
.
(1.72)
MDT MR
It includes a SU (2) triplet giving rise to the ML mass term, a Higgs doublet leading to the
Dirac mass term MD and a gauge singlet identified with MR . This seesaw mass matrix is
diagonalized by a unitary matrix U6×6 producing six mass eigenstates,
T
U6×6
M6 U6×6 = diag [mi , Mi ] ,
(1.73)
8
Note that a minimum of two right-handed neutrinos is needed in order to reproduce the observed
light neutrino masses. Here the (3 + 3) case is shown for illustration.
40
Chapter 1. Neutrinos in the Standard Model and beyond
three for the light neutrinos mi and three for the heavy leptons Mi . Explicitly, the light
neutrino mass is given in a seesaw Type-I+II form by [4]
mν = ML − MD MR−1 MDT ,
(1.74)
where the smallness of neutrino masses stems from assuming MR MD ML .
It is easy to see that for a standard Type-I mechanism, the light neutrino mass becomes
mν = −MD MR−1 MDT .
(1.75)
In order to explain briefly the origin of neutrino mass through seesaw schemes, we
may rewrite the neutrino mass matrix as [17]
Y3 v3 Yν hΦi
YνT hΦi MR
Mν =
!
,
(1.76)
where Y3 and Yν are the Yukawa coupling submatrices (complex in general), and v3 and
v2 ≡ hΦi are the Higgs triplet vev and the SM Higgs doublet vev respectively. In this
scenario, the effective light neutrino mass is calculated from the equation 9
mν = Y3 v3 − Yν MR−1 YνT hΦi2 .
1.4.2
(1.77)
Inverse seesaw
One of the most interesting properties of the seesaw mechanism is its versatility. In
fact, new important features arise when the lepton sector is extended adding extra gauge
singlets, NLc and SL [18, 19], giving rise to a (3 + 6) description. In this scheme, it
is considered three light states and six heavy states, forming three pseudo-Dirac quasidegenerate pairs. In the νL , NLc and SL basis, the seesaw mass matrix is defined as


0 MD 0


Mν =  MDT
0 M ,
0 MT µ
(1.78)
where M and µ are SU (2) singlet complex mass matrices. Note that the νL − νL and
NLc −NLc terms are neglected, as suggested in several string models [20], and lepton number
9
The naturalness of the Yukawa coupling (Y ∼ 1) would imply the need of a very massive neutral
lepton as the messenger of the seesaw.
1.4. Neutrino mass models
41
Figure 1.4: Inverse seesaw mechanism where NLc and SL are included [18].
is broken only by the nonzero µij Si Sj mass terms.
This description belongs to the so-called inverse seesaw mechanism, where the masses
of the three light neutrinos are defined, after the diagonalization [21], by
mν = MD (M T )−1 µ M −1 MDT .
(1.79)
Notice that, in contrast to the standard seesaw mechanism, the smallness of neutrino
mass emerges from the tiny mass value arising from µ 10 . In fact, the following condition
characterizes this scheme,
µ MD M .
(1.80)
This “inverse” behavior of µ induces the name inverse seesaw to this mechanism.
For the sake of clarity, again we can translate Eq. (1.78) in terms of Yukawa couplings
and vevs in order to illustrate the mass generation [4, 18] as it can be seen in Fig. 1.4.
We obtain


0
Yν hΦi 0


Mν =  YνT hΦi
(1.81)
0
M 
0
MT
µ
for the inverse seesaw mass matrix, Eq. (1.78). From this mass matrix, the light neutrino
mass is expressed as
mν = hΦi2 Yν (M T )−1 µ M −1 YνT .
(1.82)
10
The neutrino mass disappears when µ → 0.
42
Chapter 1. Neutrinos in the Standard Model and beyond
1.4.3
Linear seesaw
An alternative description of the seesaw mechanism arises from a SO(10) unified model [22].
It is based on the same basis as inverse seesaw, νL , NLc and SL , but the seesaw mass matrix
obtained after the rupture of the extended gauge structure is


0 MD ML


Mν =  MDT
0 MR  ,
MLT MRT 0
(1.83)
giving rise to the following effective neutrino mass:
mν = MD (ML MR−1 )T + (ML MR−1 )MDT .
(1.84)
It is worth noting that if ML → 0, the three light neutrinos become massless, recovering
the SM.
Again, in order to shed light to the origin of neutrino mass, we rewrite Eq. (1.83) as


0
Yν hΦi F hχL i


Mν =  YνT hΦi
0
F̃ hχR i  ,
F T hχL i F̃ T hχR i
0
(1.85)
leading to the following light neutrino mass:
mν '
i
hΦi2 h
Yν (F F̃ −1 )T + (F F̃ −1 )YνT ,
Munif
(1.86)
where Munif indicates the unification scale and F and F̃ represent independent combinations of Yukawa coupling of SL .
One can observe that Eqs. (1.84) and (1.86) are linear in the Yuakawa couplings of
Dirac neutrinos. This is the reason why this description is called linear seesaw.
1.4.4
Left-right symmetric model
Left-right symmetric models are based on the SU (3) × SU (2)L × SU (2)R × U (1)B−L
structure and are characterized by the parity conservation of weak interactions. It implies
that each left-handed fermion must have a right-handed partner, that is, νL and NR for
the neutrino sector. This scenario generates a seesaw mass matrix as in Eq. (1.72), with
1.4. Neutrino mass models
43
the following expression in terms of Yukawa couplings and vacuum expected values:
Mν =
YL h∆L i Yν hΦi
YνT hΦi YR h∆R i
!
.
(1.87)
As in the previous models, the basis is defined by νL and NLc ; YL,R indicate the left-right
Yukawa couplings and h∆L,R i are the vevs producing the left-right Majorana mass terms.
The matrix Mν can be diagonalized using a perturbative method [17], getting the
following expression for light neutrino masses:
mν ≈ YL h∆L i − Yν YR−1 YνT
hΦi2
.
h∆R i
(1.88)
It is important to reiterate that the main idea of these models is that the smallness of
neutrino mass is induced through the exchange of a heavy singlet right-handed neutrino
(Type-I) or a heavy scalar boson (Type-II).
1.4.5
Radiative models
Neutrino masses can be also produced as a consequence of radiative corrections [23]. For
instance, in the Babu model they can appear at two-loop level [24] with this expression:
Mν ∼ λ0
1
16π 2
2
f Yl hYl f T
hΦi2
hσi .
(mk )2
(1.89)
Here, it is introduced a doubly-charged scalar k ++ that is much heavier than h+ , l indicates
a charged lepton, f , h and Yl represent the Yukawa couplings and hσi is a vev of a
SU (3) × SU (2) × U (1) singlet [25].
More details can be found in Refs. [23–25].
1.4.6
Supersymmetry as origin of neutrino mass
Low-energy supersymmetry can generate neutrino masses [26] by means of the R-parity
symmetry rupture, which produces lepton number violation. This could occur spontaneously thanks to an SU (3) × SU (2) × U (1) singlet sneutrino, which yields a non zero vev
[27–29]. The general expression related with this kind of models is given by the following
approximation:
1
A
Mν ∼
hΦi2 Yd Yd .
(1.90)
2
16π
m0
44
Chapter 1. Neutrinos in the Standard Model and beyond
For an extensive description of supersymmetry as origin of neutrino mass, please see
Refs. [30, 31].
1.5
Neutrino oscillations in vacuum
Neutrino oscillations are the most sensitive evidence of neutrino mass and, therefore, that
physics beyond the Standard Model is needed. The idea of neutrino oscillations is very
simple and was introduced by Pontecorvo in 1957 [32, 33].
Neutrinos are produced as flavor eigenstates as a result of charged-current weak interactions. The neutrino hamiltonian in this basis is not diagonal. This fact makes a
difference between the flavor states, νe , νµ and ντ , and the mass states, ν1 , ν2 and ν3 .
Thus, since the flavor states are a combination of the mass states (Eq. (1.92)), the probability of finding a neutrino with the same flavor state that was created oscillates with
time, because the mass states evolve as
|νi (t)i = e−iEi t |νi (0)i .
(1.91)
This phenomenon is called neutrino oscillations.
For their detection, let us consider, for instance, a pion beam decaying into a muon
and a muon-neutrino (π ± → µ± + νµ (ν̄µ )). Although only muon-neutrinos are produced
initially, some electrons can be registered in the detector. This result is obtained because some muon-neutrinos have oscillated to electron-neutrinos producing electrons via
charged-current weak interactions.
A similar phenomenon was detected by the Super-Kamiokande Collaboration in June
of 1998 [1] when reported the muon-neutrino disappearance (νµ → ντ ) from their atmospheric neutrino data, providing the first evidence of neutrino oscillations [34]. In subsequent years, several experiments have evidenced the neutrino oscillation phenomenon
as SNO [35, 36], which measured the solar neutral neutrino flux through neutral and
charged-current weak interactions, establishing the conversion of electron-neutrinos into
muon and tau-neutrinos, or KamLAND [37] that confirmed neutrino oscillations at the
solar sector. Finally, the oscillations of atmospheric neutrinos were also confirmed at the
long baseline accelerator experiments as K2K [38], T2K [39–41] and MINOS [42–44].
1.5. Neutrino oscillations in vacuum
1.5.1
45
The neutrino oscillation probability in vacuum
Let us start considering that there are n flavor eigenstates, να , which are connected with
n mass eigenstates νi , through the relations [10, 12]
|να i =
X
∗
Uαi
|νi i ,
|νi i =
X
Uαi |να i ,
(1.92)
α
i
where greek indices represent flavor states and latin indices mass eigensates. U is the
neutrino mixing matrix, also known as Maki-Nakagawa-Sakata (MNS) matrix for the
three-neutrino case 11 , and it satisfies the following properties:
U †U = 1 ,
X
∗
Uαi Uβi
= δαβ ,
X
∗
= δij .
Uαi Uαj
(1.93)
α
i
For the case of n active neutrinos, U is a n × n matrix characterized by n(n − 1)/2
Euler angles and n(n + 1)/2 phases. If we are considering Dirac neutrinos, the number
of physical phases is (n − 1)(n − 2)/2, being responsible of CP violation in the lepton
sector. However, if neutrinos have Majorana nature the matrix U contains n(n − 1)/2 CP
violation phases. Then, in general, the U matrix can be written as
U = V P,
(1.94)
where V is a matrix containing the Dirac phases, and P is a diagonal matrix including
(n − 1) Majorana phases (β1 , β2 , β3 , . . . , βn−1 ) in the form
P = diag(1, eiβ1 , eiβ2 , ...., eiβn−1 ) .
(1.95)
For example, with n = 3 (usual number of light active neutrinos) we find one Dirac phase
and two Majorana phases, all three associated with CP violation.
Let us assume that a flavor state |να i has been produced in t = 0. What is the
probability of finding a neutrino in a flavor state |νβ i after a time t? We should follow the
evolution of the system in order to answer this question. The initial state of the produced
neutrino is
X
∗
|ν(x, 0)i = |να i =
Uαi
|νi i ;
(1.96)
i
11
An explicit expression of this matrix is given in Sec. 1.5.3 and Refs. [7, 9, 45].
46
Chapter 1. Neutrinos in the Standard Model and beyond
and after evolving a time t,
|ν(x, t)i =
X
∗ −iEi t
Uαi
e
|νi i =
i
X
∗
Uαi
Uβi eipi x e−iEi t |νβ i ,
(1.97)
i,β
where pi is the momentum of the emitted neutrino and Eq. (1.92) has been used.
Therefore, the probability amplitude of finding a neutrino at a time t in the flavor
state |νβ i is given by
A(να → νβ ; t) = hνβ |ν(x, t)i =
X
∗
Uαi
Uβi eipi x e−iEi t .
(1.98)
i
Considering the emitted neutrino as a relativistic particle, pi mi , we can do the following approximation,
q
m2
m2
(1.99)
Ei = m2i + p2i ' pi + i ' E + i ,
2pi
2E
obtaining
A(να → νβ ; t) =
X
∗
Uαi
Uβi
i
m2 L
exp −i i
2 E
= A(να → νβ ; L) .
(1.100)
Here, it is considered t ' x = L since relativistic neutrinos travel at nearly the speed of
light, being L the distance to the detector 12 . From Eq. (1.100), we can find the expression
for the transition probability [12]:
Pνα →νβ (L, E) = |A(α → β; t)|2 =
X
∗
∗ −i(Ej −Ei )t
Uαi
Uαj Uβi Uβj
e
(1.101)
i,j
=
X
∗ 2
|Uαi
| |Uβi |2
i
+2
X
Re
∗
∗
Uαi
Uαj Uβi Uβj
i>j
∆m2ij L
exp −i
2E
,
with the mass difference defined as
∆m2ij = m2i − m2j .
(1.102)
At this point, we should do some remarks about the expression of the oscillation
probability, Eq. (1.101):
• If L 2E/∆m2ij , the oscillatory term is averaged giving rise to a constant conversion
probability.
12
Notice that we are using natural units, } = c = 1.
1.5. Neutrino oscillations in vacuum
47
• At least one mass splitting has to be different from zero, ∆m2ij 6= 0, which implies
that two neutrinos have to be massive in order to observe oscillations among three
neutrinos.
• The mixing among neutrinos is necessary in order to have oscillations, so the offdiagonal elements of U must be different from zero.
• The oscillation probability is the same if we consider Dirac neutrinos or Majorana
neutrinos. In particular, this means that we can not distinguish between Dirac and
Majorana neutrinos by studying neutrino oscillations.
• From the unitarity of the matrix U , we have that
X
P (α → β; t) =
β=e,µ,τ
X
P (ᾱ → β̄; t) = 1.
(1.103)
β=e,µ,τ
A more useful way to write the oscillation probability in Eq (1.101) is separating the
∗
∗
real and imaginary parts of the product of matrices Uαi
Uαj Uβi Uβj
, through the unitarity
condition
X
X
∗
∗
|Uαi |2 |Uβi |2 = δαβ − 2
Re Uαi
Uαj Uβi Uβj
.
(1.104)
i
i>j
From the above equation, we can rewrite the oscillation probability as
Pνα →νβ (L, E) = δαβ
∗
∆m2ij L
∗
− 2
Re Uαi Uαj Uβi Uβj 1 − cos
2E
i>j
X
∗
∆m2ij L
∗
+ 2
Im Uαi Uαj Uβi Uβj sin
.
2E
i>j
X
(1.105)
And using some simple algebra we get the most usual expression [10, 12]:
Pνα →νβ (L, E) = δαβ
∆m2ij L
− 4
Re
sin
4E
i>j
X
∗
∆m2ij L
∗
+ 2
.
Im Uαi Uαj Uβi Uβj sin
2E
i>j
X
∗
∗
Uαi
Uαj Uβi Uβj
2
(1.106)
Antineutrino oscillations
The derivation of the oscillation probability for antineutrinos is straightforward following
the same steps as for neutrinos. In analogy with the neutrino case, we can define the
48
Chapter 1. Neutrinos in the Standard Model and beyond
antineutrino flavor states as
|ν̄α i =
X
Uαi |ν̄i i .
(1.107)
i
If we use Eq. (1.107), we obtain the same expressions that in Eqs. (1.101), (1.105)
∗
∗
,
and Uβi → Uβi
or (1.106), only taking into account that Uαj → Uαj
Pν̄α →ν̄β (L, E) = δαβ
∆m2ij L
− 4
Re
sin
4E
i>j
X
∆m2ij L
∗
∗
.
+ 2
Im Uαi Uαj Uβi Uβj sin
2E
i>j
X
∗
∗
Uαi Uαj
Uβi
Uβj
2
(1.108)
CP, T and CPT implications
The CP transformation interchanges neutrinos with positive helicity to antineutrinos with
negative helicity, connecting the oscillation channels,
να → νβ ⇐⇒ ν̄α → ν̄β .
(1.109)
After this transformation we obtain Eq. (1.108) from Eq. (1.106). In general, the mixing
matrix is complex and contains phases, which could lead to CP symmetry violation. Such
violation can be measured in neutrino oscillation experiments as
ACP
αβ = Pνα →νβ − Pν̄α →ν̄β .
(1.110)
On the other hand, the T symmetry interchanges the initial and final states of a given
processes and therefore relates the oscillation channels,
να → νβ ⇐⇒ νβ → να ,
ν̄α → ν̄β ⇐⇒ ν̄β → ν̄α .
(1.111)
T symmetry violation can be also observed in neutrino oscillation experiments through
the observables
ATαβ = Pνα →νβ − Pνβ →να ,
ĀTαβ = Pν̄α →ν̄β − Pν̄β →ν̄α .
(1.112)
Finally, if we apply both transformations at the same time, we get the CPT transfor-
1.5. Neutrino oscillations in vacuum
49
mation,
να → νβ ⇐⇒ ν̄β → ν̄α ,
(1.113)
which is expected to be conserved as an underlying principle of the Standard Model:
Pνα →νβ = Pν̄β →ν̄α .
1.5.2
(1.114)
Oscillations with two neutrinos
The most simple illustration of neutrino oscillations considers only two flavor states:
νe and νµ for example, with their two corresponding mass states, ν1 and ν2 . In this
approximation, the mixing matrix U is a 2 × 2 matrix given by
U=
cos θ sin θ
− sin θ cos θ
!
,
(1.115)
where θ is the mixing angle. From Eq. (1.106), we get the following expression for the
oscillation probability in the appearance channel,
P (νe → νµ ) = P (νµ → νe ) = P (ν̄e → ν̄µ ) = P (ν̄µ → ν̄e )
∆m2 L
2
2
,
= sin 2θ sin
4E
(1.116)
whereas for the disappearance channel we have the following survival probability,
P (νe → νe ) = 1 − P (νe → νµ ) ,
(1.117)
as we can derive from Eq. (1.103).
Sometimes, it is convenient to write the probability Eq. (1.116) in the following form:
2
2
P (να → νβ ) = sin 2θ sin
∆m2 L
1.27
E
,
(1.118)
where natural units have been used (} = c = 1) and the 1.27 factor follows from the
conversion of km into eV−1 , with ∆m2 and E given in eV2 and GeV respectively.
At this point, we should do a reflexion: why is the study of the two-neutrino case
useful if we know the existence of three neutrinos at least?
• The oscillation probability of two-neutrino case is much simpler than three neutrinos.
50
Chapter 1. Neutrinos in the Standard Model and beyond
• Many experiments are not sensitive to all mass-splittings and mixing angles, so the
data can be analyzed by considering the two-neutrino approximation, as we will see
in Sec. 1.5.3.
1.5.3
Oscillations with three neutrinos
In this section, we will introduce the oscillation mechanism for three generations of neutrinos, the standard number of light active neutrinos. For this purpose, we start providing
a parameterization for the lepton mixing matrix U 3×3 , considering it unitary 13 . It is a
3 × 3 matrix which can be written, using the Okubo’s notation [46], as a product of three
complex rotations:




1
0
0
c13
0 e−iφ13 s13
c12
e−iφ12 s12 0




c23
e−iφ23 s23  
0
1
0
c12
0 ,
 0
  −eiφ12 s12
0 −eiφ23 s23
c23
−eiφ13 s13 0
c13
0
0
1
(1.119)
where cij and sij represents cos θij and sin θij respectively. Eq. (1.119) leads to the symmetrical representation of the lepton mixing matrix 14 [9, 45]


c12 c13
s12 c13 e−iφ12
s13 e−iφ13


 −s12 c23 eiφ12 − c12 s23 s13 e−i(φ23 −φ13 ) c12 c23 − s12 s23 s13 e−i(φ12 +φ23 −φ13 ) s23 c13 e−iφ23 .
s12 s23 ei(φ12 +φ23 ) − c12 c23 s13 eiφ13
−c12 s23 eiφ23 − s12 c23 s13 e−i(φ12 −φ13 )
c23 c13
(1.120)
This symmetric parameterization is equivalent to the PDG [7] description of U ,


c12 c13
s12 c13
s13 e−iδ


s23 c13  × diag(1, eiβ1 , eiβ2 ), (1.121)
 −s12 c23 − c12 s23 s13 eiδ c12 c23 − s12 s23 s13 eiδ
s12 s23 − c12 c23 s13 eiδ −c12 s23 − s12 c23 s13 eiδ c23 c13
where δ is the Dirac CP violation phase and β1 and β2 are the phases associated to
Majorana neutrinos. The connection between Eq. (1.120) and Eq. (1.121) is given by the
13
This is not a general condition since in extended models, where more than 3 neutrinos are considered,
the lepton mixing matrix involving three active neutrinos is not unitary, as we will see in chapter 5.
14
We want to emphasize this parameterization of the lepton mixing matrix U because it will be used
later in chapter 5, since it is more intuitive in order to work with more than three neutrinos.
1.5. Neutrino oscillations in vacuum
51
following relation of phases:
δ = φ13 − φ12 − φ23
β1 = φ12
(1.122)
β2 = φ12 + φ23
Although the mixing matrix has been written in Eq. (1.120) using all physical phases
φij , only one combination of them can be observed in neutrino oscillation experiments [45]:
I123 = φ12 + φ23 − φ13 = −δ .
(1.123)
Obtaining the expression of the neutrino oscillation probability in the case of three
families is straightforward from Eq. (1.106), taking into account that latin indices go from
1 to 3, and there are only two independent mass-splittings, ∆m221 and ∆m231 . With these
considerations we finally get:
P (να → νβ ) = δαβ
3
X
∆m2ij L
− 4
Re
sin
4E
i>j=1
3
X
∗
∆m2ij L
∗
+ 2
Im Uαi Uαj Uβi Uβj sin
.
2E
i>j=1
∗
∗
Uαi
Uαj Uβi Uβj
2
(1.124)
This is the most general expression of the neutrino oscillation probability in vacuum
for three generations. But in particular, we can reduce this expression to a case of twoneutrino flavors, as we commented in Sec. 1.5.2. For instance, for some atmospheric,
reactor and accelerator neutrino experiments we can assume that
∆m221
L 1,
2E
(1.125)
giving rise to a two-neutrino probability [47]:
2
2
2
Pνα →νβ (L, E) = 4|Uα3 | |Uβ3 | sin
∆m231
L
4E
.
(1.126)
For the sake of completeness, we consider another limiting case concerning solar neutrinos and long baseline reactor experiments (as KamLAND [37]), where it is assumed
that
∆m231
∆m232
L'
L 1.
(1.127)
2E
2E
52
Chapter 1. Neutrinos in the Standard Model and beyond
Figure 1.5: Feynman diagrams of neutrino interactions in matter.
In this case, the oscillations related to the large mass splittings, ∆m231 and ∆m232 , are
very fast, so they can be averaged. Thus, the survival probability presents the following
form [47]:
4
Pνe →νe (L, E) ' cos4 θ13 P2 + sin θ13
,
(1.128)
where P2 is the νe survival probability in the case of two neutrinos:
2
2
P2 = 1 − sin 2θ12 sin
1.6
∆m221
L .
4E
(1.129)
Neutrino oscillations in matter
Neutrinos propagating in matter are subject to interactions with particles in the medium.
They can be absorbed by the matter constituents, or suffer coherent and incoherent
scattering in their way through matter. These interactions are encoded in an effective
potential which depends on the density of fermions in the medium and modifies the
mixing of neutrinos, leading to an effective mixing angle in matter which can be large
even for small values of the vacuum mixing angle. Thus, matter can enhance the flavor
conversion of neutrinos, giving rise to a resonant effect under certain conditions of medium
density and neutrino energy. This is known as the Mikheyev-Smirnov-Wolfenstein (MSW)
effect [12, 48, 49] and it explains the so-called solar neutrino problem [47, 50–55]. We will
be back to this effect in the following lines.
1.6.1
Neutrino evolution in matter
In their trip through matter, neutrinos interact with fermions by means of neutral-current
weak interactions (NC) regardless of their flavor, but only electron-neutrinos do it with
1.6. Neutrino oscillations in matter
53
electrons through charged-current weak interactions (CC), as we can see in Fig. 1.5.
In the case of electron-neutrinos interacting with electrons through CC, the effective
hamiltonian has the following expression:
GF
HCC = √ [ēγµ (1 − γ5 )νe ] [ν̄e γ µ (1 − γ5 )e] ,
2
(1.130)
where GF is the Fermi constant 15 . After a Fierz transformation [11, 12, 57, 58] we
obtain:
GF
(1.131)
HCC = √ [ēγµ (1 − γ5 )e] [ν̄e γ µ (1 − γ5 )νe ] .
2
The effective potential VCC is given by
Z
VCC = hνe |
d~x HCC (~x)|νe i .
(1.132)
After solving this integral [11, 12], we get
VCC =
√
2 GF Ne ,
(1.133)
with Ne being the electron number density in the medium. Similarly, we can proceed
in order to calculate the NC contributions to the potential (VN C ), considering, in this
case, all neutrino flavors in Fig. 1.5. Again, we start with the neutral-current effective
hamiltonian,
i
X h
GF X
HN C = √
[ν̄α γ µ (1 − γ5 )να ]
f¯γµ (gVf − gAf γ5 )f ,
2 α=e,µ,τ
f =e,n,p
(1.134)
where gVf and gAf are the vector and axial coupling constants of the Standard Model
respectively. In an electrically neutral medium, contributions coming from protons and
electrons are canceled, so they have no influence on the potential VN C :
gVe
gVp
gVn
15
1
= − + 2 sin2 θW ,
2
1
=
− 2 sin2 θW ,
2
1
= − .
2
Notice that Eq. (1.130) is related to Eqs. (1.16) and (1.19) considering GF ≡
natural units.
(1.135)
√
2
8
g
MW
2
[56] in
54
Chapter 1. Neutrinos in the Standard Model and beyond
Therefore, the matter potential due to neutral-current weak interactions is given by [11, 12]
VN C = −
1√
2GF Nn ,
2
(1.136)
where Nn is the neutron number density. Then, considering all contributions, the matter
potential for each type of neutrino will be expressed as follows:
Vνe
Vνµ
√
= 2GF
Nn
= VCC + VN C
,
Ne −
2
√
Nn
= Vντ = VN C = 2GF −
.
2
(1.137)
(1.138)
Notice that, since antineutrinos have opposite isospin, we have to change Vνα → −Vν̄α .
Let us now consider the evolution of neutrinos in matter, starting with the most simple
scheme of two neutrinos. The evolution equation for neutrinos in vacuum in the flavor
basis 16 is given by
νe
νµ
d
i
dt
d
i
dt
νe
νµ
!
E1 0
0 E2
=U
!
2
− ∆m
cos 2θ
4E
∆m2
sin 2θ
4E
=
!
∆m2
4E
∆m2
4E
U†
sin 2θ
cos 2θ
νe
νµ
!
!
,
νe
νµ
(1.139)
!
.
(1.140)
In order to derive the neutrino evolution in matter we have to introduce the matter
potentials affecting νe and νµ :
d
i
dt
νe
νµ
!
2
=
− ∆m
cos 2θ + VCC
4E
∆m2
sin 2θ
4E
∆m2
4E
∆m2
4E
sin 2θ
cos 2θ
!
νe
νµ
!
.
(1.141)
Note that the neutral-current potential VN C does not appear because it is common to
electron-neutrinos and muon-neutrinos. Therefore, it does not modify the Schrödinger
equation (Eq. (1.141)) and can be absorbed 17 .
16
For the case with matter, it is more convenient to use the flavor basis because the effective potentials
are diagonal in this basis.
17
Having iψ̇ = Hψ, we can add a phase to each diagonal term of H, such that H 0 = H + α. In this
0
0
0
case, if we define ψ = e−iαt ψ we can find easily that iψ̇ = Hψ ⇐⇒ iψ̇ = Hψ.
1.6. Neutrino oscillations in matter
1.6.2
55
Oscillations in a medium of constant density
After deriving the hamiltonian in matter, further assumptions can be made in order to
simplify the study of matter effects in neutrino oscillations. For instance, we can consider
that the matter density is constant, i.e. Ne = constant.
After the diagonalization of the effective hamiltonian in matter, we find the following
mass eigenstates:
ν1m = νe cos θm + νµ sin θm
(1.142)
ν2m = −νe sin θm + νµ cos θm ,
where θm is the mixing angle in matter, related with the vacuum mixing parameters by
tan 2θm =
∆m2
2E
∆m2
2E
sin 2θ
√
.
cos 2θ − 2 GF Ne
(1.143)
Finally, the oscillation probability in a medium of constant density has the following
expression:
∆m2matt
m
2
2
L ,
(1.144)
Pνe →νµ = sin 2θm sin
4E
with the energy difference given by
E1m − E2m =
∆m2matt
2E
"
=
√
∆m
cos 2θ − 2 GF Ne
2E
2
2
+
2
∆m
2E
2
#1/2
sin2 2θ
. (1.145)
Note that the probability in matter of constant density, Eq. (1.144), is analogous to
the probability in vacuum, Eq. (1.116), considering the mixing parameters in matter,
Eqs. (1.143) and (1.145).
Let us mention a special condition in a few lines. As it is derived from the oscillation
amplitude,
2 2
∆m
sin2 2θ
2E
sin2 2θm =
,
(1.146)
√
2
2
∆m2
∆m2 2
2
G
N
+
sin
cos
2θ
−
2θ
F
e
2E
2E
a resonance effect is produced when the condition
VR ≡
√
∆m2
2 GF NeR =
cos 2θ
2E
(1.147)
is satisfied, being NeR the electron density at the resonance point and VR the resonance
56
Chapter 1. Neutrinos in the Standard Model and beyond
potential. This is known as the Mikheyev-Smirnov-Wolfenstein (MSW) effect [12, 48, 49]
and it explains the behavior of neutrino oscillations inside a material medium, as the
Earth or the Sun.
1.6.3
Oscillations in an adiabatic medium
Let us now consider a medium where the density is variable. With this condition,
Eq. (1.141) is not easy to solve and, in general, it is not possible to find an analytic
solution, so numerical techniques have to be applied. However, when the density changes
slowly (adiabatic medium), Eq. (1.141) can be worked out, obtaining an approximate
analytic solution.
For an adiabatic medium as we have described above, the mixing angle in matter θm ,
defined in Eq. (1.143), depends on the medium density, Ne . Therefore, as neutrino travels
through the adiabatic medium, θm changes, acquiring a time dependence θm (t):
νe
νµ
!
cos θm sin θm
− sin θm cos θm
=
!
ν1m
ν2m
!
ν1m
ν2m
= U (θm )
!
,
(1.148)
The evolution equation in the mass basis [11] is
i
ν̇1m
ν̇2m
!
m
= U † (θm )MW
U (θm )
ν1m
ν2m
!
ν1m
ν2m
− iU † (θm )U̇ (θm )
!
,
(1.149)
where we have used,
d
dt
νe
νµ
!
= U̇ (θm )
ν1m
ν2m
!
+ U (θm )
ν̇1m
ν̇2m
!
.
(1.150)
Writing Eq. (1.149) in a more suitable way, we obtain
i
ν̇1m
ν̇2m
!
=
E1m (t) −iθ̇m (t)
iθ̇m (t) E2m (t)
!
ν1m
ν2m
!
,
(1.151)
m
where E1,2
(t) are defined in Eq. (1.145), taking into account their dependence with the
variation of Ne in the adiabatic medium and, as a consequence, their time dependence.
As it can be observed, there is a significative difference with respect to Eq. (1.139) since
the hamiltonian is not diagonal in this case. This is due to the time dependence of θm in
a medium where the density is not constant, changing the mass basis too. Nevertheless,
1.6. Neutrino oscillations in matter
57
in an adiabatic medium we find that
|θ̇m | |E1m − E2m | ,
(1.152)
then the off-diagonal terms in the evolution equation, Eq. (1.151), are small and can be neglected. This is called the adiabatic approximation, and it can be used when the adiabatic
condition [11, 47],
∆m2
sin 2θ
1
2|θ̇m |
= 2E
|V̇CC | 1 ,
(1.153)
≡ m
m
m
γ
|E1 − E2 |
|E1 − E2m |3
is satisfied. Here, γ is the adiabaticity parameter, with |E1m − E2m | and VCC defined by
Eqs. (1.145) and (1.133) respectively.
In order to achieve a better understanding of the adiabatic case, an illustrative example
will be very useful. As starting point, we consider the usual description of neutrino mixing
in a two flavor scheme:
νe = ν1m cos θm + ν2m sin θm ,
(1.154)
νµ = −ν1m sin θm + ν2m cos θm ,
(1.155)
or their complementary expressions:
ν1m = νe cos θm − νµ sin θm ,
(1.156)
ν2m = νe sin θm + νµ cos θm .
(1.157)
Under this description, an electron-neutrino νe is produced as initial state in a medium
with very high density (the core of the Sun, for instance), which implies VCC VR . We
may redefine Eq. (1.143) as
tan 2θ
,
(1.158)
tan 2θm =
1 − VVCC
R
i
deriving that the matter mixing angle in the production point θm
is approximately equal
to π/2. Then, no neutrino mixing occurs due to the matter effect and therefore νe is a pure
ν2m . As neutrino travels through matter, the density decreases, diminishing the related
mixing angle in matter and giving rise to an increase in the neutrino mixing up its maximal
value θm = π/4, which is the value of the matter mixing angle at the resonance point
(VCC = VR ). Continuing its journey, the neutrino reaches zones with smaller density
and, as a consequence, the mixing angle in matter keeps decreasing until a minimum
f
identified with the vacuum mixing angle θm
= θ, when VCC VR . Summarizing this
58
Chapter 1. Neutrinos in the Standard Model and beyond
example, we can say that the electron-neutrino νe , produced initially as ν2m , evolves to
νe sin θ0 + νµ cos θ0 during its propagation.
As it can be observed, there is no transition between the mass eigenstates ν2m and
ν1m (the neutrino state remains as ν2m ). On the other hand, a variation in the flavor
composition is produced as neutrino propagates in matter because θm depends on the
matter density, leading to the following transition probability between the two flavor
eigenstates:
P (νe → νµ ) = cos2 θ.
(1.159)
In a more general case, where no initial conditions are defined, we find the following
expression for a newborn neutrino:
i m
i m
ν(ti ) = νe = cos θm
ν1 + sin θm
ν2 .
(1.160)
After propagation in matter, the final neutrino state becomes [47]
i −i
ν(tf ) = cos θm
e
R tf
ti
E1m (t0 )dt0 m
ν1
i −i
+ sin θm
e
R tf
ti
E2m (t0 )dt0 m
ν2
(1.161)
at a time tf . Keeping in mind the time dependence (due to density dependence) of θm ,
we find that the transition probability is expressed as
Pνe →νµ =
1
1 1
i
f
i
f
− cos 2θm
cos 2θm
− sin 2θm
sin 2θm
cos Φ ,
2 2
2
with
Z
tf
Φ=
(E1m − E2m )dt0 .
(1.162)
(1.163)
ti
f
Notice that, if the high density initial condition and θm
= θ final condition are included
in Eq. (1.162), the transition probability recovers the form of Eq. (1.159).
1.6.4
Oscillations in a non-adiabatic medium
As a final case, we will analyze the most generic case, when the adiabaticity condition
is not fulfilled and the medium density does not vary softly. In this framework, where
|θ̇m | ∼ |E1m − E2m |, transitions between the eigenstates ν1m and ν2m are allowed due to
the violation of adiabaticity, giving rise to non-zero off-diagonal terms in Eq. (1.151).
Under these requirements, the conversion probability in a non-adiabatic medium is given
1.6. Neutrino oscillations in matter
59
∆χ
2
10
5
0
0.2
0.3
0.4
0.4
2
0.6
0.01
0.02
sin θ12
sin θ13
sin θ23
∆χ
2
10
0.03
2
2
5
0
6
7
2
∆m21
2
8
-5
2
∆m31
2
[10 eV ]
2.5
-3
0
2
[10 eV ]
0.5
1
1.5
2
δ/π
Figure 1.6: Plots stemming from the χ2 analysis for the neutrino oscillation parameters [59].
by [11, 47]
1 1
i
f
− cos 2θm
cos 2θm
(1 − 2PLZ ) ,
(1.164)
2 2
where PLZ is the transition probability between ν1m and ν2m , known as Landau-Zener
probability [11]. This probability is expressed as
P (νe → νµ ) '
π PLZ ' exp − γr .
2
(1.165)
In the above expression, γr stands for the adiabaticity parameter at the resonant point,
given by the MSW condition Eq. (1.147), and it is presented as follows [11, 47]:
sin2 2θ ∆m2
γr =
cos 2θ 2E
VCC V̇ .
CC res
(1.166)
In the adiabatic limit, γr 1, when the Landau-Zener probability vanishes, the probability defined in Eq. (1.162) is recovered.
60
Chapter 1. Neutrinos in the Standard Model and beyond
Table 1.3: Neutrino oscillation parameters from a global fit performed after Neutrino-2014
conference [59].
Best fit ± 1σ
2σ range
3σ range
7.60+0.19
−0.18
7.26 − 7.99
7.11 − 8.18
∆231 [10−3 eV2 ] (NH)
+0.05
2.48−0.07
2.35 − 2.59
2.30 − 2.65
∆231 [10−3 eV2 ] (IH)
+0.05
2.38−0.06
2.26 − 2.48
2.20 − 2.54
3.23 ± 0.16
2.92 − 3.57
2.78 − 3.75
5.67+0.32
−1.15
4.17 − 6.22
3.95 − 6.42
5.73+0.25
−0.38
4.38 − 6.20
4.05 − 6.39
2.10+0.14
−0.09
1.90 − 2.35
1.79 − 2.47
2.16+0.10
−0.12
1.93 − 2.38
1.82 − 2.50
δ/π (NH)
1.48+0.43
−0.39
0.00 − 0.31 & 0.72 − 2.00
0.00 − 2.00
δ/π (IH)
1.48+0.28
−0.29
0.00 − 0.04 & 0.89 − 2.00
0.00 − 2.00
Parameter
∆221 [10−5 eV2 ]
sin2 θ12 /10−1
sin2 θ23 /10−1 (NH)
2
−1
sin θ23 /10
(IH)
sin2 θ13 /10−2 (NH)
2
−2
sin θ13 /10
(IH)
To finish this section, it is worth noting that the analysis performed above could be
completed introducing a third neutrino. We can find works with a three-neutrino scheme
in Refs. [11, 47], for example.
1.7
Oscillation data
In order to close this review of neutrino features in the Standard Model and beyond, we
will summarize neutrino oscillation data coming from neutrino experiments. In particular,
we will use [59] as reference 18 , which includes an update of neutrino oscillation parameters
performed after the Neutrino-2014 conference [62]. Specifically, a global analysis was done
using the last data of Double Chooz [62, 63], Daya Bay [64–66] and RENO [66, 67] reactor
experiments, together with T2K [39–41] and MINOS [42–44] results. The latest solar data
from Super-Kamiokande (SK in what follows) [68, 69] and SNO [70, 71], as well as older
solar data coming from Homestake [50], Gallex/GNO [72], SAGE [73], Borexino [74],
SK-I [75] and SK-II [76] experiments are also included.
18
We can find other compilations of neutrino oscillations data in Refs. [60] and [61], for instance.
1.7. Oscillation data
61
The value of the most relevant neutrino oscillation parameters are compiled in Fig. 5.9
and Table 1.3. Notice that NH stands for Normal Hierarchy, where ∆m231 > 0, and IH
indicates Inverted Hierarchy, with ∆m231 < 0.
62
Chapter 1. Neutrinos in the Standard Model and beyond
Chapter 2
Non-Standard Interactions
The Standard Model is one of the most successful theories in particle physics, being
confirmed in most of the existing experiments of high energy physics. However, the
SM introduces neutrinos as massless particles (Sec. 1.2.3), contradicting the neutrino
oscillation measurements [59]. Hence, the SM is not the end of the road and needs to
be revised in order to introduce neutrino masses and mixings, giving rise to new physics
beyond the Standard Model. Most of these models, which include mechanisms of neutrino
mass generation (such as seesaw schemes, described in Sec. 1.4), imply extra interactions
called Non-Standard Interactions (NSI). So far, there is no experimental evidence of the
existence of interactions beyond those described within the framework of the SM plus
oscillations, although their study is interesting from a phenomenological point of view,
since their presence might indicate what type of physics beyond the Standard Model is
favored.
These new interactions were used to explain the solar [50–54] and the atmospheric
anomalies [1, 77–82], since NSI could induce neutrino conversions, as a resonant effect,
even in the absence of the neutrino mass [83]. After the results of the KamLAND experiment it became clear that NSI can only play a sub-leading role [2, 84–86] and that
neutrino oscillations is the main mechanism to explain neutrino data. Indeed, there exist
in the literature several theoretical and phenomenological analyses of NSI involving accelerator, reactor, solar or supernova neutrino experiments. Combining data from all of
these experiments it is possible to obtain bounds on NSI parameters, as we will see later
in chapters 3 and 4.
63
64
Chapter 2. Non-Standard Interactions
Table 2.1: Coupling constants of fermions with the neutral-current weak boson Z.
NC couplings
gLf
νe , νµ , ντ
1
2
0
e, µ, τ
− 21 + sin2 θW
1
− 23 sin2 θW
2
− 12 + 32 sin2 θW
2
u, c, t
d, s, b
2.1
gRf
sin θW
− 32 sin2 θW
1
3
sin2 θW
Parameterization of NSI
At low-energy scales, the Standard Model interactions involving neutrinos are described
by a four-fermion effective lagrangian:
X f
√
√
Lef f = −2 2 GF [ν̄α γρ L lα ][f¯γ ρ Lf 0 ] + h.c. − 2 2 GF
gP [ν̄α γρ Lνα ][f¯γ ρ P f ] , (2.1)
P,f,α
where GF is the Fermi constant, P = L, R correspond to the chirality operator, L =
(1 − γ 5 )/2 and R = (1 − γ 5 )/2, l is a charged lepton, f is a fermion, with f 0 as its SU (2)
partner and gPf are the weak neutral-current couplings given in Table 2.1. The first term
of Eq. (2.1) stands for the charged-current weak interaction mediated by the W boson,
while the second one corresponds to the neutral-current weak interaction with Z as the
gauge boson (Fig. 1.5).
If NSI are taken into account, they are introduced in theory as extra terms in the
effective four-fermion lagrangian, where the strength of these extra interactions is given
by the new couplings ε [87–90]:
X f,P
√
SI
εαβ [ν̄α γρ Lνβ ][f¯γ ρ P f ] ,
LN
N C = −2 2 GF
(2.2)
P,f
where now, f is a first generation fermion (e, u or d). Notice that Eq. (2.2) is the
neutral-current non-standard interaction lagrangian, in which this thesis will focus. For
completeness, we describe the effective lagrangian involving charged-current non-standard
interactions, which presents a similar structure to Eq. (2.1):
X CC,f,f 0 ,P
√
0
SI
LN
=
−2
2
G
εαβ
[ν̄α γρ L lβ ][f¯γ ρ P f ] .
F
CC
0
f,f ,P
(2.3)
2.2.
2.2
NSI from extended models
65
NSI from extended models
NSI could appear in several extensions of SM as a consequence of the extra interactions
following from neutrino mass as, for instance, in seesaw models (Sec. 1.4). In addition,
extended models which include extra particles (additional fermions, neutral heavy leptons,
Higgs triplet, etc) imply new physics interactions which can be parameterized using the
NSI formalism described by the Eqs. (2.2) and (2.3). In order to illustrate how the NSI
formalism parameterizes these models beyond SM, we will describe several of them in this
section.
2.2.1
Models with extra neutral gauge bosons
Some extended models introduce extra gauge bosons in addition to those arising from
the SM gauge symmetry SU (2)L × U (1)Y . For instance, a neutral massive boson Z 0 is
predicted in E6 string-inspired models, coming from extra groups U (1)χ and U (1)ψ . In
fact, one may take Z 0 as a combination of bosons of both groups with a β mixing angle.
Another example is the left-right symmetric model, which is based on the gauge group
SU (2)L × SU (2)R × U (1)B−L implying an expanded fermion sector (with right-lepton
doublets) and new gauge bosons, Z 0 and W 0 for neutral-current and charged-current
interactions respectively.
The existence of extra bosons introduces new interactions added to the standard ones.
As an example, in a four-fermion description of the neutrino-quark scattering, the lagrangian is represented by the following expression:
µ
qL µ
GF X 5
5
qR
C
5
√
ν̄
γ
(1
−
γ
)ν
f
q̄γ
(1
−
γ
)q
+
f
q̄γ
(1
+
γ
)q
,
LN
=
−
e
µ
e
νq
2 q=u,d
(2.4)
where f qP are the SM coupling constants [6]
f
uL
=
f dL =
f uR =
f dR =
1 2
2
− κ̂νN ŝZ + λuL + εuL ,
2 3
1 1
NC
2
ρνN − + κ̂νN ŝZ + λdL + εdL ,
2 3
2
NC
2
ρνN − κ̂νN ŝZ + λuR + εuR ,
3
1
NC
2
ρνN
κ̂νN ŝZ + λdR + εdR .
3
C
ρN
νN
(2.5)
(2.6)
(2.7)
(2.8)
66
Chapter 2. Non-Standard Interactions
In these expressions, where we consider the M S scheme [6], ŝ2Z stands for sin2 θW (with
C
uL
dL
dR
θW as the Weinberg mixing angle) and ρN
= 2λuR are the radiative
νN , κ̂νN , λ , λ , λ
corrections, whose values are given in [6]. The new interactions are included as εqP
parameters which can be identified as NSI couplings. In particular for E6 string inspired
models, these NSI parameters are expressed as [91]
εuL =
MZ2
−4 2
MZ 0
C
sin2 θW ρN
νN
r !
cos β sin β 5
√ −
3
8
24
r !
3 cos β sin β 5
√
+
,
6
8
2 24
(2.9)
εdR = −8
MZ2
C
sin2 θW ρN
νN
MZ2 0
r !2
3 cos β sin β 5
√
+
,
6
8
2 24
(2.10)
εdL = εuL = −εuR ,
(2.11)
with MZ as the mass of the standard neutral gauge boson Z and MZ 0 denotes the mass
of the heavier extra gauge boson Z 0 , stemming from the combination of U (1)χ and U (1)ψ
gauge bosons.
From the lagrangian in Eq. (2.4) we may get the differential cross section of neutrinonucleus scattering,
dσ
=
dT
G2F M
2π
"
(GV + GA )2 + (GV − GA )2 1 −
T
Eν
2
− (G2V − G2A )
MT
Eν2
#
,
(2.12)
where M represents the nucleus mass, T is the recoil energy of the nucleus (going from
0 to Tmax = 2Eν2 /(M + 2Eν )) and Eν is the energy of the incident neutrino. The NSI
contribution can be found in the vectorial and axial parameters GV and GA given by [91]
GV
GA
V
p
dV
2
dV
n
uV
(2.13)
(gV + 2εuV
ee + εee ) Z + (gV + εee + 2εee ) N Fnucl (Q ) ,
p
dA
A
2
dA
n
uA
= (gA + 2εuA
ee + εee )(Z+ − Z− ) + (gA + εee + 2εee )(N+ − N− ) Fnucl (Q ) .
=
(2.14)
Here, Z and N denote the number of protons and neutrons respectively, with a ± subscript
V,A
indicating spin-up and spin-down, and Fnucl
(Q2 ) are the vector and axial nuclear form
factors. As can be seen from Eqs. (2.13) and (2.14), the NSI contribution is introduced
through εqV,A
parameters, but using the vectorial-axial description 1 .
ee
1
Vectorial-axial NSI parameters are related with those coming from a left-right description through
2.2.
NSI from extended models
67
For the sake of completeness, we show the influence of NSI in the neutrino-electron
scattering characterized by the following lagrangian,
GF X C
LN
ν̄α γµ (1 − γ 5 )νβ f eL ēγ µ (1 − γ 5 )e + f eR ēγ µ (1 + γ 5 )e ,
νe = − √
2 α,β=e,µ,τ
(2.15)
e
with f eL,R identified with the coupling constants gL,R
but including the NSI corrections
eL,R
ε
[91]:
e
f eL,R = gL,R
± εeL,R ,
εeL = 2
MZ2
C
sin2 θW ρN
νe
2
MZ 0
εeR = 2
MZ2
C
sin2 θW ρN
νe
MZ2 0
(2.16)
r !2
3 cos β sin β 5
√ +
,
(2.17)
3
8
2 6
r !
r !
cos β sin β 5
3 cos β sin β 5
√ −
√
+
.
3
8
3
8
2 6
24
(2.18)
Again, we may define the differential cross section as
"
#
2
2G2F me
m
T
dσ
T
e
=
(gLe )2 + (gRe )2 1 −
− gLe gRe
,
dT
π
Eν
Eν2
(2.19)
e
,
with NSI parameters correcting the SM coupling constants gL,R
1
gLe = − + sin2 θW + εeL ,
2
e
gR = sin2 θW + εeR .
2.2.2
(2.20)
(2.21)
Seesaw models
As we have seen in Sec. 1.4, seesaw models [9, 15, 92, 93] introduce extra neutral heavy
leptons that it may imply a rich phenomenology to study. In these schemes, extra isosinglets are added to the standard SU (2)L × U (1)Y isodoublets giving rise to a possible
mixing between them [9, 94]. This mixing is described by the rectangular matrix 2
K = (KL , KH ) ,
(2.22)
1
1
qV
qV
qA
qA
the following relations: εqL
and εqR
ee = 2 εee + εee
ee = 2 εee − εee .
2
Notice that this matrix comes from the truncation of the matrix U n×n in a model with n > 3
neutrinos, where it is only considered the active ones.
68
Chapter 2. Non-Standard Interactions
where KL describes the mixing between the standard light neutrinos, while KH includes
the mixture with neutral heavy leptons. In this framework, we can find new physics
hints from this matrix as deviations from the standard charged-current interactions as a
consequence of extra couplings in the lagrangian, or in neutral-current processes because
KL is no longer unitary 3 . Indeed, neutral-current interactions are described in the SM
by the following lagrangian:
ig 0
Zµ ν̄L γµ U † U νL
2 sin θW
ig 0
Zµ ν̄L γµ νL ,
=
2 sin θW
L =
(2.23)
where we have used the unitarity condition of U . After adding extra neutral heavy leptons
the lagrangian presents a significant change:
L=
ig 0
Zµ ν̄L γµ K † KνL .
2 sin θW
(2.24)
Here, K † K 6= I leading to the spontaneous appearance of NSI [9, 94]. These new interactions can be introduced in the theory using the ε parameters. For instance, considering
only one extra heavy neutrino 4 , NSI parameters could be expressed as
2
εeL
ee = −gL sin θ14 ,
2
εeR
ee = −gR sin θ14 ,
(2.25)
where θ14 is the mixing between the light neutrino and the neutral heavy lepton, and gL,R
are the corresponding SM couplings.
On the other hand, certain seesaw models imply an extended Higgs sector too. In
particular, for a Type-II seesaw model a Higgs triplet is introduced in order to provide
mass to neutrinos, as we can observe in Fig. 2.1. In this case, the NSI parameters are
related with the neutrino mass matrix as follows [95]:
m2∆
√
ερσ
=
−
(mν )σβ (m†ν )αρ ,
αβ
8 2GF v 4 λ2φ
(2.26)
where v denotes the vev of the SM Higgs, m∆ stands for the Higgs triplet mass and λφ is
related to the Higgs triplet coupling.
3
We will explain this case more accurately in Chapter 5.
One extra heavy neutrino implies a global extra factor in the effective lagrangian of electron-neutrinoelectron scattering as a consequence of the non-unitarity of KL [9, 94].
4
2.2.
NSI from extended models
69
Figure 2.1: Diagrams for the exchange of a Higgs triplet in a Type-II seesaw model at
tree-level [95].
2.2.3
SUSY with R-parity violation
In supersymmetric theories [96], the baryon number B and the lepton number L can be
violated, implying that R-parity 5 is broken [97–99]. In the low-energy regime, these
supersymmetric models may introduce NSI arising from extra violating L terms in the
superpotential [30, 31, 100–102]
λijk Li Lj Ekc ,
0
λijk Li Qj Dkc ,
(2.27)
(2.28)
where L and Q represent the super-fields containing the standard lepton and quark doublets, while E c and Dc denote the super-fields including singlets, and i, j, k are generational
indices.
In this kind of models, the R-violating interactions produced by the heavy supersymmetric particles can be described by a four-fermion effective lagrangian involving leptons
and quarks. In the case where neutrinos interact with d-quarks we have:
X
√
µ
¯ µ
εdR
Lef f = −2 2GF
αβ ν̄Lα γ νLβ dR γ dR .
(2.29)
α,β
5
R-parity is defined as R = (−1)3B+L+2S , where B and L indicates the baryon and lepton number
respectively, whereas that S stands for the spin.
70
Chapter 2. Non-Standard Interactions
In the above expression the NSI parameters are present, representing the strength of extra
interactions. As an example, we can identify flavor-conserving NSI parameters [86, 103],
0
εdR
µµ
=
X
j
|λ2j1 |2
√
,
4 2GF m2q̃jL
(2.30)
0
εdR
ττ
=
X
j
|λ3j1 |2
√
,
4 2GF m2q̃jL
(2.31)
and flavor-changing NSI couplings,
0
εdR
µτ
=
X
j
0
λ λ
√ 2j1 3j12 .
4 2GF mq̃jL
(2.32)
In these equations, mq̄L indicates the masses of the squarks, identifying the subscripts
j = 1, 2, 3 as d˜L , s̃L , b̃L respectively. These neutral NSI couplings lead to extra nonuniversal and flavor-changing terms relevant to solar [84, 85, 104, 105] and atmospheric
propagation [86, 106, 107], as we will show in chapter 3.
2.2.4
Models with leptoquarks
In several extensions of the SM an exotic boson appears coupling indistinctly to a lepton
and a quark: the leptoquark [108]. For example, the Pati-Salam model [109], grand unification theories based on SU (5) [110–112] and SO(10) groups [113] or extended technicolor
models [114].
Due to the presence of these leptoquarks, NSI must be considered along with standard
interactions, producing extra contributions to the four-fermion lagrangian given by the
following NSI parameters [91, 115]:
ε
uV
εdV
√
λ2u
2
=
,
2
mlq 4GF
√
λ2d
2
=
,
2
mlq 4GF
(2.33)
(2.34)
where λu,d are coupling constants and mlq represents the leptoquark mass. We should
note that Eqs. (2.33) and (2.34) are valid for vector leptoquarks. An extra 1/2 factor has
to be added for the case of scalar leptoquarks [115].
2.3. NSI phenomenology
2.3
71
NSI phenomenology
Studying NSI can be very attractive from a phenomenological point of view, since its
existence would mean a new evidence of new physics. The effects of these new interactions could be observed in the production (εS ), propagation (εm ) and detection (εD ) of
neutrinos. In this section we will show how NSI influence each of these processes. Later
we will introduce these new interactions in the data analysis of several kinds of neutrino
experiments (Sec. 2.4).
2.3.1
NSI in the source and detector
Production and detection are charged-current processes, so extra contributions as in
Eqs. (2.1) and (2.3) could be introduced in the presence of NSI, giving rise to an extra flavor component in both processes [89, 116–119]:
|ναS i = |να i +
X
εSαρ |νρ i = (1 + εS )U |νi i ,
(2.35)
†
D †
εD
γβ hνβ | = hνi |U [1 + (ε ) ] ,
(2.36)
ρ=e,µ,τ
hνβD |
= hνβ | +
X
γ=e,µ,τ
where the superscripts S and D indicate “source” and “detector” respectively, and |νi i is a
mass eigenstate. Notice that the orthonormality between |ναS i and hνβD | is broken, because
NSI matrices, εS and εD , are arbitrary and non-unitary (in general) due to different
physical processes (related with NSI) that can take place at the source and at the detector
of the experiment. Therefore, it is possible to detect a flavor transition at zero distance [83,
94, 120], as a consequence of the broken unitarity. Using Eqs (1.101), (2.35) and (2.36)
we find that the transition probability is
2
X
L
m2
i
∗ −i 2E PναS →νβD (L) = (1 + εD )γβ (1 + εS )αρ Uρi Uγi
e
γ,ρ,i
X
X
∆m2ij L
j∗
j∗
i
i
2
=
Jαβ Jαβ − 4
Re[Jαβ Jαβ ] sin
4E
i,j
i>j
X
∆m2ij L
j∗
i
+ 2
Im[Jαβ Jαβ ] sin
,
2E
i>j
(2.37)
72
Chapter 2. Non-Standard Interactions
i
contains the terms related with NSI,
where Jαβ
i
∗
Jαβ
= Uαi
Uβi +
X
∗
εSαγ Uγi
Uβi +
X
γ
∗
εD
γβ Uαi Uγi +
γ
X
∗
εSαγ εD
ρβ Uγi Uρi .
(2.38)
γ,ρ
As we can observe in Eq. (2.37), the first term is independent of the distance (L), so the
flavor conversion could be produced at the source, i.e., at zero distance.
2.3.2
Neutrino propagation with NSI
On the journey of neutrinos through the Earth, NSI can contribute to coherent forward
scattering processes in matter, which are neutral-current interactions parameterized by
Eq. (2.2). Indeed, the presence of these NSI would modify the neutrino matter potential,
adding a new term to the standard oscillation hamiltonian (Eq. (1.141)):



f
f
f
ε
ε
δ
0
m21 0
ef + εee
eτ
eµ
 † X 
 1 
f  , (2.39)
f ∗
f
2
U
+
V
= U
ε
ε
ε
0
0
m


f 
µτ 
eµ
µµ
2
2E
f
f ∗
f ∗
2
εµτ ετ τ
0
0 m3
εeτ

Hmat+N SI

where U is the lepton mixing matrix for three neutrinos (Eqs. (1.120) and (1.121)). The
elements of the NSI matrix, εαβ , are related with those in Eq. (2.2) as
εαβ =
X Vf
f,P
Ve
εf,P
αβ ,
(2.40)
representing the strength of NSI, mi is the neutrino mass value, E stands for the neutrino
√
energy, Vf = 2GF Nf denotes the interaction potential with fermions (f = e, u or d),
where Nf is the fermion number density along the neutrino path throughout the mat√
ter. Note that Ve = 2GF Ne is the standard potential of neutrinos with matter, which
produces the usual MSW effect (Eq. (1.133)).
Using this hamiltonian (Eq. (2.39)) we can write the evolution equation including NSI
as

 
νe

 
i dtd  νµ  = U
ντ

1
2E


 

0
0
0
δef + εfee εfeµ εfeτ
νe
P



 

∗
0  U † + f Vf  εfeµ
εfµµ εfµτ   νµ .
 0 ∆m221
∗
∗
ντ
0
0
∆m231
εfeτ
εfµτ ετ τ
(2.41)
Note that neutrinos, in their propagation, are only sensitive to vector currents
6
We will return to this property in Chap 3.
6
(εV =
2.3. NSI phenomenology
73
εL + εR ), thus Eq. (2.40) may be written in a left-right notation as:
εαβ =
X Vf
f
Ve
f,R
(εf,L
αβ + εαβ ) .
(2.42)
Equation (2.41) provides the structure for the neutrino propagation in matter when
NSI are present. Analyzing this expression we may find some interesting details concerning
NSI couplings and their influence in neutrino propagation:
• The diagonal elements of NSI matrix, εαα , could induce extra resonance oscillation effects, additional to those following from the MSW matter potential, even if
neutrinos were massless particles [83, 121].
• The off-diagonal elements, εαβ , could produce flavor transitions in matter, even if
no neutrino mixing is present [83, 121, 122].
The observation of these phenomena would be considered an evidence for NSI and a signal
of new physics.
In a similar procedure to the vacuum case, Sec. 1.5.1, we can diagonalize the hamiltonian Hmat+N SI using a unitary matrix Ũ ,

H̃ = Hmat+N SI

m̃21 0
0
1 
 †
= Ũ
 0 m̃22 0  Ũ ,
2E
0
0 m̃23
(2.43)
with m̃2i as the effective neutrino mass-squared eigenvalues and Ũ is the effective mixing
matrix when matter and NSI are taken into account.
The transition probability is the square of the probability amplitude, hence
2
3
X
m̃j L
∗
−i 2E Pνα →νβ (L, E) = Ũαj Ũβj e
.
(2.44)
j=1
So we get, after applying the same calculation as in Sec. 1.5.1,
Pνα →νβ (L, E) = δαβ
∆m̃2ij L
− 4
4E
i>j
X
∆m̃2ij L
∗
∗
+ 2
Im[Ũαi Ũαj Ũβi Ũβj ] sin
.
2E
i>j
X
∗
∗
Ũαj Ũβi Ũβj
] sin2
Re[Ũαi
(2.45)
74
Chapter 2. Non-Standard Interactions
Notice that Eq. (2.45) looks like the neutrino transition probability in vacuum, Eq. (1.124),
but replacing the vacuum masses m2i and the leptonic mixing matrix U , for those effective
parameters which includes the interaction with matter (Ve ) and NSI, m̃2i and Ũ respectively. Therefore, in order to study the new physics behind NSI in propagation, it is useful
to compare vacuum and effective parameters and relate them [123]. The effective masses
with NSI have the following expressions:
m̃21 ' ∆m231 Ṽ + αs212 + Ṽ εee ,
h
i
2
2
2
2
∗
m̃2 ' ∆m31 αc12 − Ṽ s23 (εµµ − ετ τ ) − Ṽ s23 c23 (εµτ + εµτ ) + Ṽ εµµ ,
i
h
m̃23 ' ∆m231 1 + Ṽ ετ τ + Ṽ s223 (εµµ − ετ τ ) + Ṽ s23 c23 (εµτ + ε∗µτ ) ,
(2.46)
(2.47)
(2.48)
whereas the elements of the effective mixing matrix are given by
αs12 c12
+ c23 εeµ − s23 εeτ ,
Ṽ
s13 e−iδ + Ṽ (s23 εeµ + c23 εeτ )
'
,
1 − Ṽ
' c23 + s223 c23 Ṽ (ετ τ − εµµ ) + s23 Ṽ (s23 εµτ − c223 ε∗µτ ) ,
' s23 + Ṽ c23 εµτ + s23 c223 (εµµ − ετ τ ) − s223 c23 (εµτ + ε∗µτ ) .
Ũe2 '
(2.49)
Ũe3
(2.50)
Ũµ2
Ũµ3
(2.51)
(2.52)
Here, sij = sin θij , cij = cos θij , α ≡ ∆m221 /∆m231 , Ṽ ≡ Ve /∆m231 and δ stands for the CP
violation phase.
It is worth noting that neutrino data is consistent with the oscillation mechanism.
Thus, NSI will play only a sub-leading role in interpreting the results of neutrino experiments. NSI couplings have been constrained using solar and atmospheric neutrino
oscillation experiments (Secs. 2.4.3 and 2.4.1 respectively), combining these data with
those coming from laboratory experiments 7 [84–87, 124]. On the other hand, reactor
neutrinos do not have relevance in constraining NSI in the propagation because the matter effect is negligible in this case. However, they can constrain NSI related with detection
processes [125].
Two-neutrino propagation with NSI
As we have seen in Sec. 2.3.2, the expressions involving three-flavor neutrino propagation
in matter including NSI, Eqs. (2.41) and (2.45), are quite complicated. Sometimes working
7
We will present detailed constraint calculations in Chapter 3 and 4.
2.3. NSI phenomenology
75
with a two-neutrino scheme is enough to illustrate the important features of NSI physics
and simplifies tedious calculations. Hence, in order to shed light on neutrino oscillations
with NSI, we will develop the conversion equations for the two-neutrino scheme (electron
and tau-neutrino in this case, because muon-neutrino bounds are stronger) step by step 8 .
The hamiltonian for a two-neutrino scheme is given as a combination of vacuum,
matter and NSI hamiltonian,
H̃ = Hvac + Hmat + HN SI ,
(2.53)
where
Hvac
Hmat
HN SI
!
−∆m2 cos 2θ ∆m2 sin 2θ
1
=
4E
∆m2 sin 2θ ∆m2 cos 2θ
!
√
VCC
2 GF Ne 0
2
=
⇐⇒
0
0
0
!
√
εee εeτ
=
2 GF Nf
⇐⇒
εeτ ετ τ
,
(2.54)
!
0
,
− VCC
2
0 ε
0
ε ε
(2.55)
!
,
(2.56)
with θ as the two-neutrino mixing angle, Ne and Nf represents the electron and fermion
density respectively along neutrino trajectory in matter, GF is the Fermi constant, VCC =
√
2GF Ne indicates the standard charged-current weak interaction of neutrinos with elec√
trons, and the strength of NSI is represented by the coupling constants, ε = 2GF Nf εeτ
√
0
and ε = 2GF Nf (ετ τ − εee ). With all these ingredients, we can build the complete
hamiltonian H̃,
∆m2
sin 2θ + ε
4E
0
∆m2
cos 2θ + ε
4E
2
H̃
=
cos 2θ + VCC
− ∆m
4E
∆m2
sin 2θ + ε
4E
2
⇔
− ∆m
cos 2θ +
4E
∆m2
4E
VCC
2
sin 2θ + ε
−
0
∆m2
4E
ε
2
∆m2
4E
!
(2.57)
!
sin 2θ + ε
cos 2θ −
VCC
2
0
+
ε
2
.
(2.58)
Note that, in Eqs. (2.55), (2.56) and (2.57), we take again into account that the Schrödinger
equation does not change if a common phase is introduced in the diagonal terms of the
hamiltonian. Using this property, we can write the hamiltonian of Eq. (2.58) in a sym-
8
We can find an accurate discussion about the two-neutrino scheme in presence of NSI in Ref. [126].
76
Chapter 2. Non-Standard Interactions
metrical way, leading to the following evolution equation:
d
i
dt
νe
ντ
!
=
−a b
b a
!
!
νe
ντ
,
(2.59)
with
0
∆m2
VCC ε
a =
cos 2θ −
+ ,
4E
2
2
∆m2
sin 2θ + ε .
b =
4E
(2.60)
(2.61)
From Eqs. (2.59), (2.60) and (2.61) we may obtain the parameters and equations that
govern neutrino conversion in the presence of NSI. Specifically, the effective mixing angle
ϕ in a two-neutrino scheme, propagating in matter with NSI is given by
2b
tan 2ϕ =
=
a − (−a)
∆m2
4E
∆m2
4E
sin 2θ + ε
cos 2θ −
VCC
2
+
ε0
2
,
(2.62)
and the transition probabilities are
Pνe →νe (L, E) = 1 − Pνe →ντ (L, E) ,
(2.63)
Pνe →ντ (L, E) = sin2 2ϕ sin2 (ωL) ,
(2.64)
with the oscillation parameter ω 2 = a2 + b2 . This parameter characterizes the neutrino
transition, presenting the following form:
2
ω =
0
VCC ε
∆m2
cos 2θ −
+
4E
2
2
2
+
∆m2
sin 2θ + ε
4E
2
.
(2.65)
Notice that, if we consider antineutrinos instead of neutrinos, we should introduce −Ne
and −Nf for the charged-current and neutral-current potential respectively, changing the
sign of interactions with matter and NSI.
2.4
NSI and neutrino experiments
In this section we will review the impact of NSI in atmospheric, solar or laboratory
neutrino experiments.
2.4. NSI and neutrino experiments
2.4.1
77
NSI in atmospheric neutrino experiments
The analysis of atmospheric neutrinos can be performed in a two-neutrino scheme, considering the non-universal νµ + f → νµ + f and flavor changing νµ + f → ντ + f processes,
so that all the NSI parameters related with electron-neutrino, εeα , are set to zero. In this
approximation, the evolution equation is
d
i
dt
νµ
ντ
!
"
= U
0
0
0
∆m231
2E
!
Vf ε
0
ε ε
U† +
!#
νµ
ντ
!
,
(2.66)
√
√
0
noting that ε = 2 GF Nf εµτ and ε = 2 GF Nf (ετ τ − εµµ ). Assuming that neutrinos
only interact with d quark and ∆m221 L/E → 0, the survival probability is given by
[86, 89, 90, 106, 127–129]
2
2
Pνµ →νµ (L, E) = 1 − Pνµ →ντ (L, E) ' 1 − sin 2ϕ sin
∆m231 L
R ,
4E
(2.67)
where the effective parameters ϕ and R for the two-neutrino approximation with NSI are
expressed as
1 2
2
2
sin
2θ
+
R
sin
2ξ
+
2R
sin
2θ
sin
2ξ
,
0
0
R2
1
R = 1 + R02 + 2R0 (cos 2θ cos 2ξ + sin 2θ sin 2ξ) 2 .
sin2 2ϕ =
(2.68)
(2.69)
The NSI couplings are introduced in the following parameters:
s
4E
ε0 2
2+
,
|ε|
∆m231
4
1
2ε
.
ξ =
arctan
2
ε0
R0 =
(2.70)
(2.71)
Using the Super-Kamiokande (SK in what follows) experimental data within a twoneutrino framework, the SK Collaboration obtained the following bounds on NSI couplings
at 90 % C.L. [128]:
|εµτ | < 1.1 × 10−2
and |ετ τ − εµµ | < 4.9 × 10−2 .
(2.72)
In a more recent analysis, considering IceCube-79 and DeepCore data [130], a more re-
78
Chapter 2. Non-Standard Interactions
strictive constraints have been found at 90 % C.L.:
|εµτ | < 6.0 × 10−3
and |ετ τ − εµµ | < 3.0 × 10−2 .
(2.73)
It is worth noting that NSI parameters involving electron-neutrinos cannot be constrained directly from atmospheric data, but they are related with atmospheric NSI parameters through the expression [131],
ετ τ ∼
2.4.2
3|εeτ |2
.
1 + 3εee
(2.74)
NSI in accelerator neutrino experiments
In accelerator neutrino experiments the main channels are νµ → νµ and νµ̄ → νµ̄ , so
the relevant NSI parameters will be εµµ , εµτ and ετ τ . For instance, from MINOS experiment [132–136] data, the NSI parameter εµτ has been constrained, assuming again two
neutrinos and taking the survival probability as
2
∆m
V
31
L .
− εµτ
Pνµ →νµ (L, E) ' 1 − sin2 4E
2E (2.75)
The bound obtained on εµτ at 90 % C.L. from the MINOS Collaboration is [137]
− 0.20 < εµτ < 0.07 .
2.4.3
(2.76)
NSI in solar neutrino experiments
There are several studies on NSI involving solar neutrinos [104, 138–145]. Collaborations
such as Super-Kamiokande, SNO or Borexino [139, 145] have used their data in order to
provide bounds on the NSI parameters. A recent updated work on solar NSI constraints
is given in Ref. [146], where the two-neutrino approximation is again made [147],
d
i
dr
νe0
νµ0
!
"
= Uθ12
0
0
0
∆m221
2E
!
Uθ†12 +
X
f
c213 δef − εfD εfN
∗
εfN
εfD
!#
νe0
νµ0
!
(2.77)
with νe0 and νµ0 coming from

 

 
 

νe0
c12 s12 0
ν1
c12 s12 0
νe
 0  

 
 †

 νµ  =  −s12 c12 0   ν2  =  −s12 c12 0  U  νµ  .
ντ0
0
0 1
ν3
0
0 1
ντ
(2.78)
2.4. NSI and neutrino experiments
79
Table 2.2: Bounds on NSI parameters at 90 % C.L. for the interaction of neutrinos with
d-quark, following from oscillation experiments (propagation). Only one parameter is
considered at a time [90].
NSI parameter
Bound
Reference
εdee − εdµµ
(0.02, 0.51)
[146]
εdττ − εdµµ
(−0.01, 0.03)
[146]
εdττ − εdµµ
(−0.049, 0.049)
[128]
εdττ
(−0.036, 0.031)
[130]
εdeµ
(−0.09, 0.04)
[146]
εdµτ
εdµτ
εdµτ
εdeτ
εdeτ (with εdee = −0.5)
εdeτ (with εdee = 0.5)
(−0.01, 0.01)
[146]
(−0.011, 0.011)
[128]
(−0.0061, 0.0056)
[130]
(−0.13, 0.14)
[146]
(−0.05, 0.05)
[128]
(−0.19, 0.13)
[128]
−
εdµµ
The NSI parameters εD and εN are expressed as [146]
εD = c13 s13 Re eiδCP (s23 εeµ + s23 εeτ ) − (1 + s213 )c23 s23 Re[εµτ ]
c2
s2 − s213 c223
− 13 (εee − εµµ ) + 23
(ετ τ − εµµ ) ,
(2.79)
2
2
εN = c13 (c23 εeµ − s23 εeτ ) + s13 e−iδCP s223 εµτ − c223 ε∗µτ + c23 s23 (ετ τ − εµµ ) .
(2.80)
Note that, although we are working with a two-neutrino scheme, mixing parameters involving the third neutrino are present as c13 = cos θ13 , s13 = sin θ13 or the CP violation
phase δCP , characteristic of the three-neutrino setup. In addition, NSI parameters involving tau-neutrinos (ετ τ , εeτ and ετ µ ) are also present in Eqs. (2.79) and (2.80) as a third
neutrino effect.
Thus, taking into account the approximation given by Eq. (2.77), the disappearance
probability is calculated as follows,
Pνe →νe ' s413 + c413 P2 ,
(2.81)
where P2 is the survival electron-neutrino probability, Pνe0 →νe0 , arising from Eq. (2.77).
80
Chapter 2. Non-Standard Interactions
Table 2.3: Values of NSI constraints at 90 % C.L. for the neutrino-electron scattering
considering one parameter at a time [90].
NSI
One parameter at a time
εeL
ee
(−0.021, 0.052) [140]
εeR
ee
(−0.07, 0.08) [148]
(−0.08, 0.09) [90]
εeL
µµ
(−0.03, 0.03) [87]
(−0.03, 0.03) [103]
εeR
µµ
(−0.03, 0.03) [87]
(−0.03, 0.03) [103]
εeL
ττ
(−0.16, 0.11) [140]
(−0.46, 0.24) [103]
εeR
ττ
(−0.25, 0.43) [103]
εeL
eµ
(−0.13, 0.13) [103]
εeR
eµ
(−0.19, 0.19) [148]
(−0.13, 0.13) [103]
εeL
eτ
(−0.40, 0.40) [87]
(−0.33, 0.33) [103]
εeR
eτ
(−0.28, −0.05) and (0.05, 0.28) [103]
(−0.19, −0.19) [148]
εeL
µτ
(−0.10, 0.10) [87]
(−0.10, 0.10) [103]
εeR
µτ
(−0.10, 0.10) [87]
(−0.10, 0.10) [103]
Considering NSI with d-quarks only, the bounds on NSI parameters are estimated in [146]
at 90 % C.L.
− 0.25 < εD < −0.02
and
− 0.14 < εN < 0.12 .
(2.82)
The bounds on the NSI propagation parameters are compiled in Table 2.2. In the next
section we discuss short baseline neutrino experiments, based at reactor and accelerators.
2.4.4
NSI in short baseline experiments
Up to now, we have considered experiments where the neutrino propagation is the most
significant part of the experiment, but short baseline experiments, where the effects of
propagation are not very important, have also been used to probe neutrino interactions
with quarks and leptons. In this case, bounds on NSI couplings arise from comparing the
experimental cross section with the SM cross section for the interaction of neutrinos with
the corresponding lepton or quark. Note that such experiments are sensitive to the axial
2.4. NSI and neutrino experiments
81
Table 2.4: Values of NSI constraints at 90 % C.L. for the neutrino-electron scattering
considering two parameters at a time [90].
NSI
Two parameters
εeL
ee
(−0.02, 0.09) [149]
(−0.036, 0.063) [150]
εeR
ee
(−0.11, 0.05) [149]
(−0.10, 0.09) [90]
εeL
µµ
(−0.033, 0.055) [103]
εeR
µµ
(−0.040, 0.053) [103]
εeL
ττ
(−0.51, 0.34) [149]
(−0.16, 0.11) [140]
εeR
ττ
(−0.35, 0.50) [149]
(−0.40, 0.60) [103]
εeL
eµ
(−0.53, 0.53) [151]
εeR
eµ
(−0.53, 0.53) [151]
εeL
eτ
(−0.53, 0.53) [151]
εeR
eτ
(−0.53, 0.53) [151]
εeL
µτ
(−0.53, 0.53) [151]
εeR
µτ
(−0.53, 0.53) [151]
coupling 9 , unlike propagation experiments.
As an example, we will consider the phenomenology related with short baseline experiments where electron-antineutrinos scatter off electrons. This is a neutral-current process,
so corrections following from Eq. (2.2) must be introduced in the interaction cross section
in order to study NSI. Therefore, the differential cross section including standard and new
physics terms is written as
"
#
2
X
X
dσ
2G2F me
Te
eR
eR 2
e
eL 2
eL 2
e
=
(gR + εee ) +
|εαe | + (gL + εee ) +
|εαe |
1−
dTe
π
Eν
α6=e
α6=e
#
"
X
Te
eR
e
eL
eR
eL
e
|εαe ||εαe | me 2
.
(2.83)
− (gR + εee )(gL + εee )
Eν
α6=e
In this expression, me stands for the electron mass, Ee represents the electron energy,
Te = Ee − me is the electron recoil energy and gLe and gRe are the SM coupling constants,
whose values can be found in Table 2.1.
Several experiments imply incoming electron-(anti)neutrino fluxes, as reactor exper9
Translating to NSI couplings, this means εA = εL − εR .
82
Chapter 2. Non-Standard Interactions
iments, where we find TEXONO [152], MUNU [153], Rovno [154], Krasnoyarsk [155],
Irvine [156], or accelerator experiments as LSND [157] and LAMPF [158]. An individual
or combined analysis of their data can be performed in order to constrain NSI parameters.
The results of these analyses are summarized in Tables 2.3 and 2.4.
Likewise, we may proceed for muon-(anti)neutrino scattering from CHARM-II data
[159]. However, the tau-neutrino case presents a greater difficulty because there is no
artificial source of these particles. Even so, it is possible study their interactions if one
considers that tau-neutrinos take part in the process e+ e− → ν ν̄γ probed at LEP [151],
or if it is a component of the solar neutrino flux [140]. Keeping these considerations in
mind, we can find some bounds for muon and tau-neutrino NSI parameters, which are
compiled in Tables 2.3 and 2.4.
In addition to the (anti)neutrino scattering off electrons, it is possible to study NSI
with quarks. For this purpose, the data of accelerator experiments such as CHARM [160],
CDHS [161] or NuTeV [162] have been analyzed, leading to bounds on NSI parameters for
processes of neutrinos with d-quarks. It is worth noting that the NuTeV Collaboration
found a discrepancy between the cross section measured by the experiment and the SM
value. After this mismatch, new reinterpretations of NuTeV results were made [163, 164],
reconciling the NuTeV results with SM predictions 10 . NSI constraints following from
these experiments will be given in Chapters 3 and 4.
10
We will go back to this puzzle in Chapter 4.
Chapter 3
Robustness of solar neutrino
oscillations
The neutrino oscillation mechanism is the correct scenario to explain solar neutrino data,
as we can infer from solar neutrino experiments [35, 36, 50, 70, 73, 75, 165–172], combining
them with reactor data from the KamLAND experiment [37]. Solar neutrino experiments
are affected by matter effects in a crucial way [173, 174], and the combination of both,
solar and KamLAND data, determines the so-called Large Mixing Angle (LMA) solution,
which has been considered as the correct explanation of data. It is however crucial to
investigate whether this solution is robust against possible additional effects inside the
Sun, for instance:
• Noise density fluctuations that are originated in the radiative zone by random magnetic fields [175–182].
• Spin-flavor precession [183, 184] that could be originated by magnetic fields in the
convective region [185, 186].
For all these examples, the KamLAND experiment plays a decisive role, implying that
any non-standard effect can only be considered as secondary [2].
Nevertheless, the possible existence of non-standard interactions still introduces an
important exception to the robustness of the solar neutrino oscillation interpretation [84,
104, 187]. In this context, it has been found that NSI may produce a degeneracy in the
solar neutrino parameter space, giving rise to the so-called LMA-Dark solution [84].
Given the precision expected in upcoming oscillation studies [188], one needs to scrutinize further the possible role of NSI [189]. As already mentioned in Chapter 2, NSI
appears naturally in neutrino mass theories [4] and are a powerful phenomenological tool
83
84
Chapter 3. Robustness of solar neutrino oscillations
to probe neutrino models. For example, NSI might give a guidance to distinguish between
the simplest high-scale seesaw models [9, 17, 92, 190, 191] and other scenarios based on
low scales, for instance, the inverse seesaw [18, 192] and the linear seesaw mechanisms [22],
or radiative models of neutrino mass [23, 24, 108].
In order to study the robustness of the oscillation interpretation of solar neutrino
data in the presence of non-standard interactions, we use data from Homestake [50],
SAGE [73], GALLEX [165, 166], Super-Kamiokande I [75, 167], SNO [35, 36, 70, 168, 169]
and Borexino [170, 171], in combination with the KamLAND reactor experiment [37].
We have used the solar fluxes and uncertainties from the updated Standard Solar Model
(SSM) [193]. It has been found that the LMA solution becomes degenerate with a new
“dark-side” solution in the presence of NSI [84]. So in this thesis, we have used new
data such as SNO phase III [70], and the measurements of Borexino [170, 171] in order
to further constrain and possibly lift the degeneracy. We have found however that the
LMA-D solution still remains.
Besides the study of solar data, we also analyze data from the CHARM accelerator
experiment [194]. The combination of both analyses plays a significant role in constraining
the non-standard neutrino d-quark interactions. Although CHARM is sensitive only to
electron-neutrino interactions, when combined with solar and KamLAND data, it allows
to improve the restrictions on the tau-neutrino non-standard axial and vector couplings.
In our analysis of non-standard interactions we parametrize them in terms of the
effective low-energy neutral-current operator for four fermions, as we have seen in Sec. 2.1:
√
LN SI = −εfαβP 2 2GF (ν̄α γµ Lνβ ) f¯γ µ P f .
(3.1)
Here, f is a first-generation fermion (e, u, d) and P = L, R are the chiral projectors. The
parameters εfαβP account for the NSI couplings of a neutrino with α and β flavor and the
left or right-handed component of a fermion f .
In this chapter, for simplicity we will consider f to be the down-type quark. It is
plausible to extend this work to include also the coupling with up-type quarks. However,
the computations are demanding. In the case of the interaction with electrons, this has
been considered before in the literature [140, 195, 196]. In this part of the thesis, we also
confine our discussion to NSI couplings of electron and tau-neutrinos, while the analysis
for the muon-neutrino case will be covered in Chapter 4. This approach is motivated
by the constraints on the νµ interactions, that are stronger than those for electron and
tau-neutrinos, as discussed before in Refs. [87, 151, 197] and will be reanalyzed in the
next chapter. Therefore, in the following analysis we will consider εdP
αµ = 0.
3.1. Solar and KamLAND restrictions on NSI
85
We will discuss in this chapter the impact of non-standard interactions in the neutrino
propagation through matter as well as in their detection. The original neutrino fluxes
are less likely to be modified by NSI, especially if we consider them as originating from
new physics in the neutral-current interactions. For the case of solar and KamLAND
neutrino experiments, we will obtain constraints for the vectorial NSI couplings 1 . This
is the natural description for an analysis involving neutrino propagation since there is no
sensitivity to the axial couplings in this case 2 . For the analysis of NSI in detectors (we
will discuss this analysis in the case of SNO, for example) we will also obtain constraints
for the axial coupling 3 .
3.1
Solar and KamLAND restrictions on NSI
In this section, we carry out a detailed study of the robustness of the standard oscillation
solution to the solar neutrino problem. To keep the number of parameters under control,
we consider a two-neutrino picture, which is adequate to illustrate our results.
3.1.1
The solar and KamLAND data
In order to analyze the effect of non-standard interactions over the LMA solution, we
include in our analysis the results from the radiochemical experiments as Homestake [50],
SAGE [73] and GALLEX/GNO [165, 166], the zenith-spectra data set from Super-Kamiokande
I [75, 167], as well as the results from the three phases of the SNO experiment [35, 36, 70,
168, 169] 4 , and the measurement of the 7 Be solar neutrino ratio reported by the Borexino
Collaboration [170, 171].
The latest version of the Solar Standard Model (SSM in what follows) [193] is used
in this analysis as well, which implies an improved determination of the neutrino flux
uncertainties, mainly thanks to the better accuracy on the 3 He-4 He cross section measurement and to the reduced systematic uncertainties in the determination of the surface
composition of the Sun. There are two different solar models introduced in Ref. [193],
that correspond to two different measurements of solar metal abundances. Our results
are similar for both models and, for definiteness we show in this thesis the higher solar
metallicity model, referred as BPS08(GS).
1
dL
dR
Remember that εdV
αβ = εαβ + εαβ .
We should point that SNO is sensitive to the NSI axial coupling as we will see in Sec. 3.1.4.
3
dL
dR
Remember that εdA
αβ = εαβ − εαβ .
4
In the third phase of the SNO experiment, 3 He proportional counters were used in order to measure
the neutral-current signal of the solar neutrino flux.
2
86
Chapter 3. Robustness of solar neutrino oscillations
Solar neutrino data are sensitive mainly to the neutrino oscillation parameters ∆m212
and sin2 θ12 . A reactor experiment which is sensitive to the same sector of neutrino parameters is the KamLAND experiment. Hence, one can combine the results of both solar
and KamLAND experiments, resulting in the Fig. 3.1. The KamLAND experiment measured the reactor antineutrino flux disappearance. It has an average baseline of 180 km
and matter effects are nearly negligible because the antineutrinos travel over the most superficial layers of the Earth, parameterizing the terrestrial crust with a constant density
profile (∼ 2.6 g·cm−3 ). This implies that the restrictions on NSI from the KamLAND
experiment will be minimal, but it will be useful to restrict the standard oscillation parameters, as can be seen in Fig. 3.1. We have analyzed the results of a total exposure
of 2881 ton·yr [37], considering the energy window above 2.6 MeV in order to avoid the
geo-neutrino background.
3.1.2
Effects of NSI in neutrino propagation
The effects of non-standard interactions on neutrino propagation can be studied analyzing
the hamiltonian describing solar neutrino evolution in the presence of NSI. These new
interactions, in addition to the standard oscillation term,
2
Hmat =
cos 2θ +
− ∆m
4E
2
∆m
sin 2θ
4E
√
2 GF Ne
∆m2
4E
∆m2
4E
sin 2θ
cos 2θ
!
,
(3.2)
add an extra term HN SI , accounting for an effective potential induced by the NSI with
matter. It may be written as:
HN SI =
√
2 GF Nd
0 ε
ε ε0
!
,
(3.3)
where ε and ε0 are the NSI effective vectorial parameters. For the limit εfαµP ∼ 0, they can
be defined as 5
dV
ε = − sin θ23 εdV
ε0 = sin2 θ23 εdV
(3.4)
eτ ,
τ τ − εee .
The quantity Nd in Eq. (3.3) is the number density of the down-type quark along the
neutrino path and θ23 is the atmospheric neutrino mixing angle. It is important to remark
again that the neutrino propagation inside the Sun or the Earth is sensitive only to the
dL
dR
vectorial component of NSI, εdV
αβ = εαβ + εαβ .
5
The reader can find the calculation of this approximation in Appendix A.
3.1. Solar and KamLAND restrictions on NSI
87
Before introducing the numerical analysis of solar neutrino data, it is convenient to
discuss the analytical formulas involving neutrino survival probabilities in the constant
matter density case. In the two-neutrino picture where neutrinos evolve in an adiabatic
regime, the survival probability can be approximated by Parke’s formula 6 [198],
P (νe → νe ) =
1
[1 + cos 2θ cos 2θm ] ,
2
(3.5)
where θ is the mixing angle in the two–neutrino scheme and θm is the effective mixing
angle in the neutrino production point inside of the Sun. If we do not include the nonstandard neutrino-matter interaction, the effective mixing angle may be calculated from
the following expression:
√
∆m2 cos 2θ − 2 2 EGF Ne
cos 2θm = q
,
√
2
∆m2 cos 2θ − 2 2 EGF Ne + (∆m2 sin 2θ)2
(3.6)
where Ne is the number density of electrons along the neutrino path and E is the neutrino
energy.
To explain the deficit of the neutrino signal in detectors, the neutrino survival probability should satisfy Pee < 0.5 which implies, according to Eqs. (3.5) and (3.6), cos 2θ > 0.
Hence, only vacuum mixing angles between 0 < θ < π4 are able to solve the solar neutrino
problem, as we can see in Fig. 3.1.
However, this scenario changes in the presence of non-standard interactions. From
Eqs. (3.2) and (3.3) one sees that the solar neutrino mixing angle in the presence of NSI
is given by the following expression:
√
∆m2 cos 2θ − 2 2 EGF (Ne − ε0 Nd )
,
cos 2θm =
[∆m2 ]matter
(3.7)
where
h
i2
√
2 2
2
0
∆m matter = ∆m cos 2θ − 2 2 EGF (Ne − ε Nd )
h
i2
√
2
+ ∆m sin 2θ + 4 2 ε EGF Nd .
(3.8)
(3.9)
As a result of the presence of non-standard parameters ε and ε0 into the equations, we
6
Notice that this is a particular case of Eq. (1.162) discussed in Sec. 1.6.3.
Chapter 3. Robustness of solar neutrino oscillations
-4
10
★
★
★
2
2
∆mSOL [eV ]
88
-5
10
0
0.2
0.4
0.6
0.8
10
2
sin θSOL
0.2
0.4
0.6
0.8
1
2
sin θSOL
Figure 3.1: Regions at 90%, 95%, 99% and 99.73% C.L. for neutrino oscillation parameters. The left panel shows separately the result from the solar (empty regions) and from
the KamLAND data (colored regions). The right panel shows the result for the combined
analysis [84].
can obtain Pee < 0.5 even for cos 2θ < 0, provided that
√
2
2 EGF Ne + ∆m2 | cos 2θ|
√
.
ε0 >
2 2 EGF Nd
(3.10)
This makes possible to explain the solar neutrino data with values of the vacuum mixing
angle larger than π4 (in the so-called “dark side”), for large enough values of ε0 . For
instance, for neutrino energies and matter densities in the typical range of the solar
regime one has ε0 & 0.6.
This degeneracy between the non-universal coupling ε0 and the neutrino mixing angle
θ, as a consequence of the effect of NSI on solar neutrino propagation, implies the presence
of an additional solution to the solar neutrino problem, the LMA-D solution.
After this theoretical discussion, now we turn our attention to the experimental data.
We perform a combined solar + KamLAND analysis using all the solar neutrino data
available at the time, as discussed in Sec. 3.1.1, along with the KamLAND result [37].
For the solar neutrino data, we compute the survival probabilities for a wide range of
∆m2sol , ε, and ε0 . For the neutrino mixing angle, sin2 θSOL 7 , we consider the whole range
of values (from 0 to 1), which includes the dark-side region. Regarding the statistics, for
7
We sholud remember that ∆m2sol = ∆m221 and θSOL = θ12 .
3.1. Solar and KamLAND restrictions on NSI
89
20
2
-5
8.5
8
∆mSOL [10 eV ]
∆χ
2
15
10
9
2
9.5
3σ
★
7.5
5
7
90% CL
0
0.2
0.4 0.6
2
sin θSOL
0.8
-3
2
3σ
90% CL
∆mSOL [ eV ]
10
-4
10
★
2
★
-5
10
-6
10
0
0.2
0.4
0.6
2
sin θSOL
0.8
10
5
10
∆χ
15
20
2
Figure 3.2: Regions at 90%, 95%, and 99% C.L. for neutrino oscillation parameters in the
presence of NSI. The left-bottom panel shows the result from the combined analysis of
solar and KamLAND data. The top-right panel presents the same result in more detail
for the region of interest. In this case the current analysis (shaded regions) is compared
with earlier results (empty regions) [84]. We also show the χ2 in other two panels. We
have marginalized over the NSI parameters in all panels.
the KamLAND analysis we use a Poisson statistics as in [199], while for the solar data we
employ the pull method [200] in order to fit the results using the BPS08(GS) Standard Solar Model. The main result is shown in Fig. 3.2. There, we plot the allowed regions (90%,
95% and 99% C.L.) in the solar neutrino oscillation parameter space (sin2 θSOL , ∆m2SOL )
obtained in the analysis of solar and solar + KamLAND neutrino data. To obtain this
result, a marginalization of the NSI 4-parameter χ2 (sin2 θSOL , ∆m2SOL , ε, ε0 ) analysis was
necessary. The ∆χ2 profiles, as a function of each parameter, are also shown. One can
90
Chapter 3. Robustness of solar neutrino oscillations
0.7
global best fit
BF without NSI
BF dark side
< Pee>
0.6
0.5
8
pp (all)
7
0.4
B (Borexino)
Be (Borexino)
0.3
8
0.2
1
B (SNO)
10
E [MeV]
Figure 3.3: Predicted neutrino survival probabilities for three different reference points
(details in the text). The probabilities have been averaged over the 8 B neutrino production
region. We also show, for reference, the rates for the pp neutrino flux, for the 7 Be line,
and for two different values for the 8 B neutrino flux, as reported from Borexino and from
SNO (third phase). The vertical lines indicate the experimental errors. The horizontal
lines show the energy range covered by the experiment.
see the presence of the dark-side region, which is a direct consequence of the NSI.
The best-fit points obtained for this χ2 analysis are given by the following parameter
values:
sin2 θSOL = 0.32,
∆m2SOL = 7.9 × 10−5 eV2 ,
ε = −0.15,
ε0 = −0.10.
(3.11)
We have also considered the best fit point in the absence of NSI, allowed with a ∆χ2 =
2.7:
sin2 θSOL = 0.30 ,
∆m2SOL = 7.9 × 10−5 eV2 ,
ε = 0.00 ,
ε0 = 0.00 .
(3.12)
Finally, the best fit point in the “dark-side” region of the oscillation parameters, which is
allowed with a ∆χ2 = 2.9, gives rise to the values:
sin2 θSOL = 0.70 ,
∆m2SOL = 7.9 × 10−5 eV2 ,
ε = −0.15 ,
ε0 = 0.95 .
(3.13)
In order to better understand the results obtained, we plot in Fig. 3.3 the neutrino
survival probabilities for these reference points (Eqs. (3.11), (3.12) and (3.13) labeled as
“global best fit”, “BF without NSI” and “BF dark side” respectively.
Notice from Fig. 3.3 that the different survival probability predictions are in very good
3.1. Solar and KamLAND restrictions on NSI
91
agreement with the measurements in the low-energy neutrino region (pp and 7 Be). In the
high-energy region, where the boron neutrino flux is predominant and matter effects are
more significant, the presence of NSI gives a slightly better fit to the data, in comparison to
the standard neutrino oscillation solution. This sensitivity in the high-energy region is due
to the different spectrum shape at high energy for the NSI solution. In this region, a flatter
spectrum is expected when NSI are introduced. The most important difference is found
in the intermediate-energy region, above few MeV. A precise measurement in this region
would break the degeneracy between these solutions. Therefore, a lower threshold boron
solar experiment (such as in the proposals of Hyper-Kamiokande and SNO [201, 202])
would be of great interest for this type of physics. Precise measurements of the beryllium
and pep fluxes may also help to break the degeneracy between the standard oscillation
and the dark-side solution.
3.1.3
Constraints on vectorial NSI
In the previous section, we have considered the effects of NSI in solar neutrino propagation.
We have shown how a new solution appears in the solar parameter space, in addition to
the standard light-side solution. In this subsection, we focus on the allowed region for the
vectorial NSI parameters, that we have defined for this case as ε and ε0 . For this purpose,
we can first obtain a restriction for either ε or ε0 by marginalizing our four-parameter χ2
analysis with respect to the other three-neutrino parameters. The results are shown in
Fig. 3.4, allowing to obtain the following bounds at 90% C.L.:
−0.41 < ε < 0.06 ,
−0.50 < ε0 < 0.19
&
0.89 < ε0 < 0.99 .
(3.14)
(3.15)
As these constraints show, there is no degeneracy in the flavor changing parameter
ε and we can constrain it to a single isolated region. However, we observe that for
relatively large values of ε0 , the flavor conserving parameter, there is another solution in
the “dark side” of the neutrino oscillation parameters. Thus, NSI still play a sub-leading
(but significant) role for neutrino conversion in matter, making the determination of solar
neutrino parameters, especially the solar mixing angle, still ambiguous.
Taking into account the value of the best fit for the atmospheric mixing angle in
Ref. [203], we can use the bound obtained in Eq. (3.14) and the expression of ε (Eq. (3.4))
92
Chapter 3. Robustness of solar neutrino oscillations
20
∆χ
2
15
10
5
90% C.L.
0
-1
-0.5
0
0.5
90% C.L.
1 -1
ε
-0.5
0
0.5
1
ε’
Figure 3.4: Constraints on ε and ε0 effective NSI parameters. These constraints come
from a global analysis of solar + KamLAND data. The ε parameter is constrained to one
region, while for the ε0 parameter we obtain two possible regions. One of these regions
corresponds to the light-side solution while the other one (less favored) corresponds to
the dark-side region.
in order to get a limit on the vectorial flavor changing coupling εdV
eτ at 90 % C.L.:
− 0.08 < εdV
eτ < 0.58 .
(3.16)
Before our analysis, the strongest constraint on this parameter was |εdV
eτ | < 0.5 [87]. We
can see from our result that solar neutrino data give a certain preference for positive
values of εdV
eτ , henceforth improving the lower bound.
3.1.4
NSI effects in neutrino detection
The presence of non-standard interactions can also affect the detection processes. In
particular, the axial component of the NSI could give rise to an extra contribution to
the neutral-current cross section detection at the SNO experiment, as we will see in this
subsection.
The neutral-current detection reaction at the SNO experiment is given by
ν + d → ν + p + n,
(3.17)
being proportional to gA2 , with gA as the axial coupling of the neutrino current to the
hadronic current [204]. Therefore, if there is a non-standard axial coupling, it will produce
3.1. Solar and KamLAND restrictions on NSI
93
an extra contribution in the NC signal of the SNO experiment. This non-standard axial
contribution is parameterized as [87]
φN C ∼ fB (1 + 2εA ) ,
(3.18)
where we have neglected the terms of order ε2A . In this expression, fB stands for the boron
neutrino flux and the effective axial parameter, εA , is defined as [87]:
εA =
X
dA
hPeα iN C (εuA
αα − εαα ) .
(3.19)
α=e,µ,τ
If we set to zero the non-standard axial couplings with up-type quarks we have
εA = −
X
hPeα iN C εdA
αα .
(3.20)
α=e,µ,τ
Notice that
dL
dR
εdA
αα = εαα − εαα
(3.21)
denotes the couplings entering in the effective lagrangian shown in Eq. (3.1). Thus, εA is
independent of the effective couplings ε and ε0 defined in Eq. (3.4).
In the NSI analysis from previous sections, we have assumed that there is no influence
of the axial component of NSI, i.e. εA = 0. This assumption was well justified, thanks to
the good agreement between the SNO NC measurement [70]:
+0.33
+0.36
6
−2 −1
φSNO
,
N C = 5.54 −0.31 (stat) −0.34 (syst) × 10 cm s
(3.22)
and the SSM boron flux prediction [193]:
fB = 5.94 ± 0.65 × 106 cm−2 s−1 .
(3.23)
We will now relax this hypothesis by considering the effect of the new NSI parameter εA
in the analysis.
For this case, we perform a new χ2 analysis which now includes five parameters,
χ2 (sin2 θSOL , ∆m2SOL , ε, ε0 , εA ). The results for this generalized 5-parameter analysis are
summarized in Fig. 3.5. In the left panel of this figure we show the regions at 90%, 95%
and 99% C.L. obtained from the marginalization of the full 5-parameter analysis (filled
colored regions). In the same panel we show, for comparison, the allowed areas obtained
from the 4-parameter analysis, without the εA coupling, discussed in the previous section
94
Chapter 3. Robustness of solar neutrino oscillations
20
8
∆χ
2
-5
2
∆mSOL[10 eV ]
10
10
2
★
90% CL
6
0.2
0.4
2
0.6
0.8
0
-0.4
sin θSOL
-0.2
0
0.2
0.4
εA
Figure 3.5: Combined five-parameter analysis, including axial couplings, of solar and
KamLAND data. In the case of the solar data, we have also considered the presence of
an axial NSI contribution in the SNO neutral-current channel for the detector. The left
panel shows the neutrino oscillation parameters, marginalized over of all NSI parameters,
while the right panel presents the χ2 profile for the εA parameter, after marginalizing over
all other parameters. The empty regions in the right panel show the case where the axial
coupling has been set to zero. The data prefer εA ∼ 0.
(empty regions). Both results are consistent, although, as expected, the inclusion of a
new parameter increases slightly the allowed region. Using our analysis it is also possible
to constrain the effective axial coupling εA . We show the ∆χ2 analysis for this case,
marginalized over the other four parameters, in the right panel of Fig. 3.5. It is clear
that data prefer εA ∼ 0, as expected from the good agreement between the predicted
boron flux and the NC measurement reported by SNO. From this study, we report the
constraint for εA at 90% C.L.
− 0.14 < εA < 0.06 .
(3.24)
We can attempt to translate this effective parameter into constraints for the NSI parameters that appear in the original lagrangian (Eq (3.1)), εdA
αα , through Eq. (3.20). From
this equation, we can notice that neglecting muon-neutrino NSI, the effective parameter
depends on four quantities: two averaged probabilities hPee iN C and hPeτ iN C and two NSI
dA
couplings εdA
ee and εeτ . The average values for the probabilities were reported by the SNO
3.2. Analysis including CHARM data
95
Table 3.1: Sensitivity of neutrino experiments to non-universal NSI parameters.
Data
Solar propagation
εdV
ee
εdV
ττ
X
X
Solar NC detection
KamLAND propagation
X
CHARM detection
X
εdA
ee
εdA
ττ
X
X
X
X
Collaboration [70]:
hPee iN C = 0.30 ± 0.03 ,
(3.25)
hPeτ iN C = 0.35 ± 0.02 .
(3.26)
With these values for the average probabilities, we can compute restrictions for the padA
rameters εdA
ee and εeτ . The allowed regions for this case will be presented and discussed
in the next section, where we will see that a combined analysis with CHARM laboratory
data will allow for additional constraints on the NSI couplings.
3.2
Analysis including CHARM data
Several laboratory experiments have also measured neutrino-nucleon scattering. Although
some of them are insensitive to neutrino oscillations, they will be helpful to constrain
neutrino NSI with d-quarks. In this section, we will focus on the results of the CHARM
experiment. We will obtain additional bounds from this experiment and we will combine
them with those already discussed in Sec. 3.1. Since the number of parameters involved
in the analysis increases, we have limited ourselves to the case of flavor-conserving nonstandard couplings. The relevant parameters for each experiment, in this approximation,
are shown in Table 3.1.
3.2.1
Constraints on non-universal NSI from CHARM
CHARM was an accelerator experiment carried out at CERN. This experiment measured
the neutral to charged-current cross section ratio for electron-(anti)neutrinos off quarks.
We have used the results published by the CHARM Collaboration for the νe q → νq [194],
96
Chapter 3. Robustness of solar neutrino oscillations
which is
Re =
σ(νe N → νe X) + σ(ν̄e N → ν̄X)
= (g̃Le )2 + (g̃Re )2 = 0.406 ± 0.140 .
σ(νe N → eX) + σ(ν̄e N → ēX)
(3.27)
For the couplings (g̃Le )2 and (g̃Re )2 , the most general expressions including all NSI
parameters are given by
2
(g̃Le )2 = (gLu + εuL
ee ) +
X
2
d
dL 2
|εuL
αe | + (gL + εee ) +
α6=e
2
(g̃Re )
=
(gRu
+
2
εuR
ee )
+
X
|εdL
αe | ,
(3.28)
α6=e
X
2
|εuR
αe |
+
(gRd
+
2
εdR
ee )
+
α6=e
X
|εdR
αe | .
(3.29)
α6=e
In order to render these Eqs. (3.28) and (3.29) more useful for our purpose, we introduce some simplifications:
• We neglect all flavor-changing non-standard contributions, implying that εqL
αe = 0
for α 6= e.
• The NSI parameters involving quark up are also omitted, implying that εuP
ee = 0.
In our analysis, the vectorial-axial notation is used to describe the coupling parameters, while the CHARM result is given in the left-right parametrization, so we transform
Eqs. (3.28) and (3.29) employing the following relations:
1 dV
(ε + εdA
ee ) ,
2 ee
1 dV
(ε − εdA
=
ee ) .
2 ee
εdL
=
ee
(3.30)
εdR
ee
(3.31)
With these conditions, we get our simplified expression for Eqs. (3.28) and (3.29) as
2
(g̃Le )
(g̃Re )2
2
1 dV
dA
=
+
+ (εee + εee ) ,
2
2
1 dV
u 2
d
dA
= (gR ) + gR + (εee − εee ) .
2
(gLu )2
gLd
(3.32)
(3.33)
Then, we can perform the χ2 for the CHARM data using the simplified form of the
couplings given above (Eqs. (3.32) and (3.33)), which depends on two NSI parameters,
3.2. Analysis including CHARM data
97
2
dA
εee
1
0
-1
-1
0
1
2
dV
εee
Figure 3.6: Constraints from the CHARM experiment, at 68%, 90%, 95% and 99% C.L.,
dA
on the NSI couplings εdV
ee and εee .
theo dV
dA
(εee , εdA
εdV
ee ):
ee and εee , through R
e
dA 2
R − Rtheo (εdV
ee , εee )
χ =
.
(σRe )2
2
(3.34)
Here, Re and σRe are defined by the result given in Eq. (3.27) and Rtheo corresponds to
the theoretical prediction for Re , which depends on the relevant NSI parameters εdV
ee and
dV
dA
εdA
ee by means of g̃Le and g̃Re . From this analysis, we obtain constraints in the (εee , εee )
plane. These are shown in Fig. 3.6 and in Table 3.2 at 68%, 90%, 95% and 99% C.L.
3.2.2
Constraints on NSI from a combined analysis
In the previous sections, we have analyzed the sensitivity of CHARM, solar and KamLAND experiments to non-standard interactions. But, as the information obtained is
complementary, we can combine the results from CHARM with the results from the analysis of solar and KamLAND data. This allows us to improve the constraints on the NSI
98
Chapter 3. Robustness of solar neutrino oscillations
3
6
CHARM (99%CL)
CHARM (99%CL)
2
4
1
dA
2
εττ
εττ
dV
solar + KamLAND (99%CL)
0
0
-1
-2
-1
solar + KamLAND (99%CL)
0
1
dV
εee
2
-2
-1
0
1
2
dA
εee
Figure 3.7: Constraints on the NSI parameters from a global analysis (colored regions),
at 90, 95 and 99% C.L. The dashed lines show the restrictions from CHARM data only
and solar + KamLAND. The left panel shows the limits for the vector NSI while the right
one shows the axial case.
couplings. We show the outcome of this combination in Fig. 3.7. In this plot, we see the
regions for the vector (left) and axial-vector (right) NSI couplings allowed by the global
analysis. They are also compared with the constraints coming only from the CHARM
data and that from the solar + KamLAND data.
The bounds obtained from the analysis of CHARM data are shown in Fig. 3.7 as
dA
vertical bands, after being translated into two independent bounds, εdV
ee and εee , coming
from the χ2 calculation. On the other hand, the allowed regions obtained from the solar +
KamLAND combination (diagonal bands in the figure) have been derived from the limits
on the non-universal NSI discussed in Sec. 3.1.
dV
The constraints on the vectorial NSI parameters, εdV
ee and ετ τ , arise from the allowed
region of the effective coupling ε0 in Eq. (3.14), after substituting the ε0 definition in
Eq. (3.4). The allowed region at 1σ for the mixing angle θ23 has also been used [203].
We should stress that there are two possible regions for the vector NSI parameters in
this plot. This is a consequence of the existence of two allowed regions for the neutrino
oscillation parameters when NSI are present (see Fig. 3.2, upper-right panel). The lower
region is related to the standard LMA solution while the upper one corresponds to the
dark-side solution.
We have also used the average probabilities in Eq. (3.26) to reanalyze the results for
3.3. Summary
99
Table 3.2: Limits on the vectorial and axial NSI parameters at 90% C.L. from the analysis
of CHARM data only, and from a global analysis that combines CHARM, solar, and
KamLAND data.
vectorial couplings
Global
−0.5 <
εdV
ee
< 1.2
−1.8 < εdV
τ τ < 4.4
one parameter at a time
CHARM −0.5 < εdV
ee < 1.2
Global
dV
dV
−0.2 < εdV
ee < 0.5 −1.1 < ετ τ < 0.4 & 1.6 < ετ τ < 2.2
axial couplings
Global
−1.5 < εdA
−0.4 < εdA
ee < 1.4
τ τ < 0.7
one parameter at a time
CHARM
−0.4 < εdA
ee < 1.4
Global
−0.2 < εdA
ee < 0.3
−0.2 < εdA
τ τ < 0.4
the effective NSI axial coupling εA (Eq. (3.24)). Thanks to this reanalysis we can obtain
dA
new bounds for the axial couplings εdA
ee and ετ τ . We can see that for both NSI parameters
(vector and axial), there is a degeneracy in the determination of the electron and tauneutrino NSI couplings (solar + KamLAND regions in Fig. 3.2). This degeneracy has
been restricted thanks to the addition of the CHARM data analysis.
We quote in Table 3.2 the allowed intervals at 90% C.L. that arise from the combined
dV
dA
dA
analysis for εdV
ee , ετ τ , εee and ετ τ . In the same table we also give the result of an analysis of
one parameter at a time. We show both results, for CHARM data only and for the global
analysis. We can compare these constraints with those of previous works (for instance
Ref. [87]). In our case, the combination of CHARM results with solar and KamLAND data
gives more restrictive bounds on the electron-neutrino NSI parameters. Concerning the
tau-neutrino NSI, the previously existing limits were those coming from the measurement
of the Z from LEP data and they have also been improved.
3.3
Summary
In this chapter, we have studied the role played by NSI in the explanation of solar neutrino
data. For this purpose, we have performed an analysis of neutrino interactions with downtype quarks using the most recent solar and KamLAND data. We have found that, taking
100
Chapter 3. Robustness of solar neutrino oscillations
into account NSI parameters, there is still room for large values of the NSI, allowing the
so-called dark-side region of the neutrino parameter space found in Ref. [84]. Other
secondary regions as LMA-0 and LMA-II, which were present in previous works, have
now disappeared.
At this point, we asked how to remove this degeneracy in future studies. In the case of
KamLAND, and other reactor experiments, we note that they are basically insensitive to
matter effects. Therefore, they are not important in removing the degeneracy. Hence, for
this purpose we have to consider the solar data. If we observe the shape of the neutrino
survival probabilities in Fig. 3.3, we notice that the best region to discriminate the dark
solution from the standard one is the intermediate energy. In this region, the relevant
experiments would be, for instance, Borexino [170–172], solar + KamLAND [205] or the
low-energy threshold analysis from Hyper-Kamiokande [206, 207]. Other experiments
that could remove this degeneracy may be either atmospheric or laboratory data. The
latter could rule out large NSI values. A possible future restriction from atmospheric and
laboratory data was studied in Refs. [84, 107]. Based on the results discussed above, we
conclude that the neutrino oscillation solution of solar neutrino data is still not robust
enough in the presence of non-standard interactions.
We finish this chapter by noticing that we have obtained limits on the vector and
axial non-standard neutrino interaction couplings with d-type quarks. In particular, we
have set bounds for flavor-diagonal and flavor-changing couplings, for electron and taudA
dA
dV
dV
neutrinos, εdV
ee , εeτ , ετ τ , εee and ετ τ . The combination of the CHARM experiment with
KamLAND and solar neutrino data has been very useful to obtain these constraints,
improving previous results [84, 87].
Chapter 4
Probing νµ NSI with accelerator and
atmospheric data
So far, we have concentrated on electron-neutrino NSI studies. For the study involving
muon-neutrinos, collider experiments are very interesting because they produce a wellcontrolled and clean muon-neutrino beam, with just a small contamination of electronneutrinos. Therefore, due to the precision of these experiments, it is expected that the
muon-neutrino NSI parameters (εfµβP ) will be better constrained than for the case of other
neutrino flavors. This occurs when we consider the interaction of muon-neutrinos with
electrons [103, 140]. On the other hand, for neutrinos interacting with quarks, the situation is different. In the case of the NuTeV experiment, despite the accurate measurement
of the νµ − N interaction [162], there was a discrepancy between their results and the predicted Standard Model parameters. Besides, previous experiments, such as CDHS [161]
and CHARM [160], did not have the same level of accuracy as the NuTeV result. Although these results could indicate the existence of new physics, introducing non-zero NSI
for muon-neutrinos (εfµαP ), it is important to notice that uncertainties arising from QCD
corrections may have been underestimated [208]. We will make a detailed analysis on this
point in this chapter, deriving new and reliable constraints on the NSI parameters coming
from these experiments.
Before starting the discussion of our analysis, it is worth noting that the limits on
NSI parameters, obtained from 1-loop dressing [87] of the effective four-fermion neutrino
vertex, can not be expressed rigorously without specifying a model. Indeed, they will be
highly model-dependent and a full analysis including all the diagrams needed to satisfy
gauge-invariance, leads to rather weak constraints on the flavor changing NSI parameters,
εqL,R
and εqL,R
[197].
µτ
µe
101
102
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
After the NuTeV results were reported, several works introduced new corrections leading to higher systematic uncertainties. In particular, a new estimate of the electroweak
mixing angle sin2 θW was presented, using the NuTeV measurement and the newly calculated uncertainties [163, 164]. Since the NuTeV data give the most stringent modelindependent constraints on the muon-neutrino NSI parameters (εµα ), it is important to
reanalyze their results introducing the effect of the new systematic uncertainties on the
NuTeV results. We will explain in this chapter how we made these computations. Besides, we will also review the status of the atmospheric neutrino constraints and we will
perform a combined analysis with laboratory results.
In order to obtain these bounds, we will work with the following hypothesis:
• We carry out our analysis in a two-neutrino scheme.
• Only NSI parameters involving muon-neutrino are nonzero, εfµβP 6= 0.
With these conditions, we will get relatively stringent constraints on the NSI interactions for muon-neutrinos, at the few 10−2 level, thanks mainly to the interplay between
laboratory and atmospheric data.
4.1
Neutrino-nucleon scattering and measurements
of sin2 θW
Neutrino scattering has been frequently used to probe different properties of the weak
neutral-current, as well as to perform measurements of several parameters like the electroweak mixing angle sin2 θW . It is known that, for experiments with an isoscalar target,
the uncertainties due to the QCD parton model cancel out, at least to a large extent [209].
If an interaction with an isoscalar target of up and down-type quarks is considered, isospin
invariance relates the most important neutral and charged-current contributions. As a
consequence, the ratio of both currents is related to the coupling constants gµP :
Rν =
σ(νµ N → νµ X)
= (gµL )2 + r(gµR )2 ,
σ(νµ N → µ− X)
(4.1)
Rν̄ =
σ(ν̄µ N → ν̄µ X)
1
= (gµL )2 + (gµR )2 ,
+
σ(ν̄µ N → µ X)
r
(4.2)
4.1. Neutrino-nucleon scattering and measurements of sin2 θW
103
where r corresponds to the ratio between antineutrino and neutrino charged-current cross
sections:
σ(ν̄µ N → µ+ X)
r=
.
(4.3)
σ(νµ N → µ− X)
The coupling constants have a contribution from both quarks, up and down:
(gµP )2 = (gµuP )2 + (gµdP )2 ,
(4.4)
with P = L, R.
If we consider the presence of non-standard interactions, we have to introduce the (nonuniversal and flavor-changing) NSI parameters that couple muon-neutrinos with quarks,
leading to the following expressions of the coupling constants (gµP ):
2
dL
dL 2
(g̃µL )2 = (gµuL + εuL
µµ ) + (gµ + εµµ ) +
X
2
|εuL
µα | +
=
(gµuR
+
2
εuR
µµ )
+
(gµdR
+
2
εdR
µµ )
+
X
α6=µ
2
|εdL
µα | ,
(4.5)
α6=µ
α6=µ
(g̃µR )2
X
2
|εuR
µα |
+
X
2
|εdR
µα | .
(4.6)
α6=µ
The values of Rν and Rν̄ parameters were measured by the CDHS [161] and CHARM
experiments [160].
Another important observable in the study of deep inelastic neutrino scattering is the
Paschos-Wolfenstein (PW) ratio, which is given by [210]
RP W =
σ(νµ N → νµ X) − σ(ν̄µ N → ν̄µ X)
Rν − rRν̄
=
= (gµL )2 − (gµR )2 .
σ(νµ N → µ− X) − σ(ν̄µ N → µ+ X)
1−r
(4.7)
The PW ratio depends weakly on the nucleus target hadronic structure. Besides, this
definition reduces the uncertainties coming from charm production and charm and strange
sea distributions. On the negative side, it adds an extra difficulty: the simultaneous
measurement of neutrino and antineutrino neutral-current cross sections makes necessary
the use of two separate neutrino and antineutrino beams.
The “ratio” observables (Eqs. (4.1), (4.2) and (4.7)) could be used to calculate several
neutral-current weak parameters. For instance, the NuTeV Collaboration uses these observables to report a measurement of the electroweak mixing angle, measuring the ratios
Rν and Rν̄ experimentally and transforming them into the ratio RP W . Once these ratios
are computed, they can be related with sin2 θW through [161, 210]
Rν(ν̄) =
1
5
− sin2 θW + sin4 θW (1 + r(−1) ) .
2
4
(4.8)
104
4.2
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
The NuTeV data and constraints on NSI
Using intense neutrino and antineutrino beams, the NuTeV Collaboration measured the
charged and neutral cross sections for neutrinos hitting an iron target. After separating
the neutral and charged events, they reported their results for the values of Rν and Rν̄ .
With these parameters, a numerical analysis was performed by the NuTeV Collaboration, obtaining a measurement of the left and right-handed neutral couplings to the light
quarks [162]:
(gµL )2 = 0.30005 ± 0.00137 ,
(gµR )2 = 0.03076 ± 0.00110 .
(4.9)
From these values it was possible to see a discrepancy with the Standard Model expectations [211]:
(gµL )2SM = 0.30399 ± 0.00017 ,
(gµR )2SM = 0.03001 ± 0.00002 ,
(4.10)
being almost a 3σ discrepancy in the left-handed coupling. Similarly, the collaboration
carried out a numerical fit getting a value for the electroweak mixing angle in the on-shell
scheme [162]:
sin2 θW = 0.22773 ± 0.00135(stat) ± 0.00093(syst) ,
(4.11)
which is again 3σ away from the value predicted in global precise electroweak fits [211]:
sin2 θW = 0.22292 ± 0.00082 .
(4.12)
This discrepancy between the NuTeV calculations for the left (gµL ) and right (gµR )
couplings and their SM prediction could be ascribed to the existence of a non-zero NSI
operator. Hence, it should take into account the εqP
µα parameters when the weak couplings
are expressed, as we have already exposed in Eqs. (4.5) and (4.6). These considerations
were borne in mind by the authors of Ref. [87], who obtained a positive hint for non-zero
uL
values of the left flavor-conserving NSI couplings, εdL
µµ and εµµ , as well as for the NSI
parameters which involve flavor-changing interactions, εqL;R
µτ .
However, before claiming the existence of non-standard interactions as possibly suggested by the NuTeV results, other more “conventional” corrections should be considered.
For instance, corrections from nuclear effects and next-to-leading-order corrections which
were neglected in the original NuTeV Collaboration analysis. As an example of inter-
4.2. The NuTeV data and constraints on NSI
105
NuTeV [Bentz et al.]
NuTeV [NNPDF Coll.]
NuTeV
SM [Global EW fit]
0.21
0.22
0.23
0.24
0.25
2
sin θW [on-shell]
Figure 4.1: Measurements of sin2 θW according to Refs. [162–164, 211].
pretations of the NuTeV anomaly in terms of conventional physics, the reader can see
Refs. [208, 212].
Here, we have reanalyzed the NuTeV anomaly in terms of new physics, taking into
account the new reported uncertainties from different theoretical groups [163, 164]. As
we will see, we have obtained new constraints on the neutrino NSI parameters that can
be considered as more reliable since they take into account this updated information.
One of the theoretical groups that has reanalyzed the uncertainties in the NuTeV
experiment is the NNPDF Collaboration. They have obtained better estimates of the
strange and antistrange parton distribution functions, introducing extra corrections to the
NuTeV calculated parameters. With this analysis, the NNPDF Collaboration reported a
new value for the weak mixing angle [163]:
sin2 θW = 0.2263 ± 0.0014(stat) ± 0.0009(sys) ± 0.0107(PDFs).
(4.13)
In the case of Ref. [164], authors have considered corrections from nuclear effects such
as the excess of neutrons in iron, the charge symmetry violation (which arises from the
mass differences of the up and down-type quarks) and the strangeness. Including all these
106
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
corrections, the collaboration reported a value for the electroweak mixing angle [164]:
sin2 θW = 0.2232 ± 0.0013(stat) ± 0.0024(sys) .
(4.14)
It can be seen from Eqs. (4.13) and (4.14), that the two new estimates for the electroweak mixing angle are very close to the SM prediction. Therefore, the inclusion of the
new uncertainties has helped to understand this problem and agreement is found, within
the error margin, as we can see in Fig. 4.1.
Having established that extra corrections to the NuTeV calculated parameters are
needed in order to reconcile them with the SM predictions, we will use the results from
NNPDF and Bentz et al. Collaborations [163, 164] to constrain the strength of the NSI couplings involving muon-neutrinos, εqP
µα . For this purpose, we adopt the Paschos-Wolfenstein
ratio RP W because:
• It depends very weakly on the hadronic structure of the nucleus target.
• It is largely insensitive to the uncertainties resulting from charm production and up
and down quark mass differences.
• It is only slightly affected by the charm and strange sea distribution.
After computing the prediction for the Paschos-Wolfenstein ratio for a particular value
of the NSI parameters (obtained from Eqs. (4.5) and (4.6)), we calculate the corresponding
χ2 function and fit experimental data. We consider the two different recent results already
discussed, Eqs. (4.13) and (4.14), and obtain constraints to εqP
µµ at 90% C.L. For the
NNPDF case are:
−0.017 < εdL
µµ < 0.025
&
0.84 < εdL
µµ < 0.88 ,
−0.24 < εdR
µµ < 0.088 ,
(4.15)
(4.16)
−0.72 < εuL
µµ < −0.67
&
−0.031 < εuL
µµ < 0.020 ,
(4.17)
−0.058 < εuR
µµ < 0.063
&
0.24 < εuR
µµ < 0.36 .
(4.18)
On the other hand, when the results of Ref. [164] are used in our analysis we get:
−0.005 < εdL
µµ < 0.005
&
0.86 < εdL
µµ < 0.87 ,
(4.19)
−0.17 < εdR
µµ < −0.11
&
−0.042 < εdR
µµ < 0.025 ,
(4.20)
−0.71 < εuL
µµ < 0.70
&
−0.006 < εuL
µµ < 0.006 ,
(4.21)
−0.014 < εuR
µµ < 0.016
&
0.28 < εuR
µµ < 0.31 .
(4.22)
4.2. The NuTeV data and constraints on NSI
107
1
uA
0.5
εµµ
dA
εµµ
0
-0.5
0
-1
0
0.5
dV
εµµ
1
-1
0
-0.5
0.5
uV
εµµ
Figure 4.2: The left panel shows the allowed regions for vector and axial NSI of neutrinos
with down-type quark, at 90% C.L. and 3σ. The up-type quark case is shown in the
right panel. We show the results for the error analysis of the NNPDF Collaboration [163]
(colored region) while the empty lines correspond to the error analysis performed by Bentz
et al. [164].
In both results, two allowed regions have been obtained for most of the NSI couplings,
reflecting the degeneracy shown in Fig. 4.2.
In a more general case, we increase the number of degrees of freedom implying the
simultaneous presence of left and right-handed NSI neutrino couplings and obtaining a
degenerate solution with hyperbolic shape, as we see in the Fig. 4.2. This degenerate
solution in both cases, down and up quarks, is easily understood from Eq. (4.7). In
the left panel of Fig. 4.2 we show the constraints for down-type quark NSI, while in the
right one we show the corresponding restrictions on NSI for the up-type quark. We have
assumed the presence of NSI with only one type of quark at a time. We present the
constraints in terms of the vector and axial couplings, instead of using the left and right
components. The reason for this is that we will combine these results with atmospheric
data (see Sec. 4.4), that are sensitive to the vector couplings.
For the case of flavor-changing NSI, we show the allowed regions for the chiral components in Fig. 4.3. As in the previous figure, we show here the result for the two different
updated error computations. In this case, only one plot is necessary since the constraints
for the u and d-type quarks coincide. This time, the allowed region can be interpreted as
qA
a hyperbola centered in the origin (εqV
µτ , εµτ ) = (0,0), that can be seen as a vestige of the
108
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
1
NNPDF
εµτ
qA
0.5
0
Bentz et al.
CHARM + CDHS
-0.5
-1
-1
-0.5
0
0.5
1
qV
εµτ
Figure 4.3: Allowed regions for vector and axial NSI of neutrinos with up and down-type
quark, at 90% C. L. and 3σ. We show the results for the error analysis of the NNPDF
Collaboration [163] using the colored region, while the empty lines correspond to the error
analysis performed by Bentz et al. [164]. The central ellipse shows the analysis of CHARM
and CDHS data, see discussion in Secs. 4.3 and 4.5.
two-fold degeneracy already discussed for the flavor-conserving case. It can be noticed,
qA
either from Fig. 4.3 or from Eqs. (4.5) and (4.7), that we constrain the product εqV
µτ · εµτ .
This implies that one coupling may be of order one, as long as the other remains small.
4.3
NSI in the CHARM and CDHS experiments
CHARM [161] and CDHS [160] are two other experiments that measured semileptonic
neutrino scattering cross sections. Their data are useful in a global fit, despite their
sensitivity being lower than the one reached by the NuTeV experiment. In particular,
they will play a role in the restrictions for the flavor-changing NSI. The measurements
reported by these collaborations are summarized in Table 4.1.
Using the values of Rν and Rν̄ measured by CHARM and CDHS, we perform a χ2
analysis taking into account the correlation between the neutrino and antineutrino ratios
of neutral to charged-current given by the parameter r. With these considerations, our
4.3. NSI in the CHARM and CDHS experiments
109
Table 4.1: Neutral to charged-current ratios reported by CHARM and CDHS. We also
show the SM prediction for comparison.
Experiment
Observable
Measurement
SM prediction
CDHS [161]
Rν
0.3072 ± 0.0033
0.3208
Rν̄
0.382 ± 0.016
0.381
r
0.393 ± 0.014
Rν
0.3093 ± 0.0031
0.3226
Rν̄
0.390 ± 0.014
0.371
r
0.456 ± 0.011
CHARM [160]
χ2 function in this case is given by
χ2 =
X
i
χ2i =
X j
j
2 −1
k
k
(Ri − Ri,N
SI )(σ )jk (Ri − Ri,N SI ) ,
(4.23)
j,k
where i refers to CHARM or CDHS and j, k run for Rν and Rν̄ . The results of our fit
analysis are shown in Fig. 4.3.
As we can observe in Fig. 4.3, the inclusion of CDHS and CHARM data provide
strong constraints for the flavor changing NSI parameters. However, it is important to
consider that the restriction follows from a discrepancy between the experimental and the
theoretical value of Rν and Rν̄ , as can be seen from Table 4.1. Therefore, as we mentioned
previously, these constraints should be considered as less robust than those obtained for
non-universal NSI discussed in Sec 4.1. Once we have pointed out this, for the sake of
completeness, we report the 90% C.L. bounds from the CHARM/CDHS data, combined
with the new analysis of the NuTeV data from the NNPDF (Bentz et al.) Collaboration:
− 0.023(−0.023) < εqL
µτ < 0.023(0.023),
−0.039(−0.036) < εqR
µτ < 0.039(0.036).
(4.24)
We should notice that these constraints in Eqs. (4.24) come mainly from CDHS and
CHARM data, as we see in the central region of Fig. 4.3.
110
4.4
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
Combining with atmospheric neutrino data
After obtaining the bounds from laboratory experiments, we turn our attention to atmospheric neutrino data. The experimental results are in good agreement with standard
neutrino oscillations [203, 213] and therefore, neutrino NSI with matter can only play a
sub-leading role in these data. It is interesting to consider atmospheric data in our global
analysis in order to “cut” the infinite branches of the hyperbola in Fig. 4.2. We will
obtain better constraints on the NSI neutrino couplings, also including more than one
NSI parameter at a time.
For this purpose, we consider NSI in matter in the context of atmospheric neutrinos
for flavor-changing, as well as for flavor-diagonal contributions [86, 214, 215]. The main
effect of NSI is the presence of an extra term in the hamiltonian that describes the
atmospheric neutrino propagation. This will translate, for the two-neutrino approach,
into the hamiltonian
!
qV
√
εqV
ε
µµ
µτ
,
(4.25)
HN SI = 2 GF Nq
qV
εqV
ε
µτ
ττ
qL
qR
where εqV
αβ = εαβ + εαβ and q = u, d. Notice that we omit the axial NSI parameter, since
the propagation effects are only sensitive to vectorial contributions.
Super-Kamiokande neutrino data have been analyzed under this assumption in Refs. [86,
214, 215]. But so far, no evidence of NSI has been found in the studied atmospheric data.
Therefore, one gets upper bounds on the magnitude of the NSI coupling strengths. For
our combined analysis, we include the full atmospheric SK-I and SK-II data sample [215]
resulting in the bounds:
−0.007 < εdV
µτ
< 0.007 ,
(4.26)
dV
|εdV
τ τ − εµµ | < 0.042 ,
(4.27)
at 90% C.L. and taking only one parameter at a time (1 d.o.f.). To obtain these limits,
we have considered the following assumptions:
• NSI are present only in the neutrino interaction with down-type quarks.
• The Earth composition is the one specified by the PREM model [216].
Similarly, we can rewrite the bounds on Eq. (4.27) in terms of the neutrino NSI
couplings with up-type quarks using the averaged ratio Nd /Nu = 1.028 [217]. In this case
4.4. Combining with atmospheric neutrino data
0.2
111
0.4
0.3
NNPDF
NNPDF
0.1
0.2
uA
εµµ
dA
εµµ
0.1
0
0
Bentz et al.
-0.1
-0.1
-0.2
-0.2
-0.1
-0.05
0
0.05
0.1
-0.3
-0.1
Bentz et al.
-0.05
0
0.05
0.1
uV
εµµ
dV
εµµ
Figure 4.4: The left panel shows the allowed regions for vector and axial flavor-diagonal
NSI of neutrinos with down-type quarks, at 90% C.L. and 3σ. The up-type quark case is
shown in the right panel. Neutrino data from accelerator and atmospheric experiments
have been combined to obtain this limit. Again, we show the results for the error analysis
of the NNPDF Collaboration [163] (colored region) while the empty lines correspond to
the error analysis performed by Bentz et al. [164].
we obtain:
−0.007 < εuV
µτ < 0.007 ,
(4.28)
uV
|εuV
τ τ − εµµ | < 0.043 .
(4.29)
It is possible to go one step further and combine the results coming from atmospheric
data with those from the previous accelerator analysis, reported in Secs. 4.2 and 4.3. With
this combination we constrain the hyperbolic branches in Figs. 4.2 and 4.3, obtaining the
allowed regions shown in Figs. 4.4 and 4.5 for the case of flavor-diagonal and flavorchanging NSI respectively. As in the previous section, the restrictions for flavor-changing
couplings consider NuTeV, CHARM, and CDHS results. For the flavor-diagonal case, only
the NuTeV data are considered. Again, for the combined result of NuTeV and accelerator
data, we only show the down-type quark restrictions, since the bounds for the up-type
case is nearly the same.
From Figs. 4.4 and 4.5 we can see the important role played by the atmospheric
neutrino data, since now the degeneracy in the vector NSI coupling has been removed.
The two parameter analysis reported here can be translated into a one parameter
112
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
0.1
εµτ
dA
0.05
0
-0.05
-0.1
-0.02
-0.01
0
0.01
0.02
dV
εµτ
Figure 4.5: Allowed regions, at 90% C. L. and 3σ, for the flavor-changing NSI neutrino
couplings with down-type quarks. In this case the result is dominated by CHARM +
CDHS and the atmospheric data. This time, the two reanalyses of the NuTeV data give
the same constraints. The constraints for the NSI with up-type quark are nearly the same.
constraint which, taking into account the reanalysis of Ref. [163] ([164]) at 90% C.L., is
−0.042(−0.042) < εdV
µµ < 0.042(0.042) ,
(4.30)
−0.091(−0.072) < εdA
µµ < 0.091(0.057) ,
(4.31)
−0.044(−0.044) < εuV
µµ
(4.32)
< 0.044(0.044) ,
−0.15(−0.094) < εuA
µµ < 0.18(0.140) .
(4.33)
We also report the one parameter constraints for the flavor-changing couplings. In
this case, the atmospheric neutrino analysis plays the main role in the restriction of the
vectorial couplings εqV
µτ , while CHARM and CDHS dominate the restrictions for the axial
qA
couplings εµτ . The bounds for the down-type quark are
−0.007 < εdV
µτ < 0.007 ,
−0.039 < εdA
µτ < 0.039 ,
(4.34)
−0.007 < εuV
µτ < 0.007 ,
−0.039 < εuA
µτ < 0.039 .
(4.35)
As a final comment in this section, we would like to point out that a detailed threeneutrino oscillation analysis combined with non-standard interactions will introduce the
4.5. Summary
113
Table 4.2: Summary of the constraints for the NSI parameters from the combined analysis
of NuTeV and atmospheric data. The first rows show the constraints for the non-universal
(NU) NSI parameters. The bottom row refers to the flavor-changing (FC) NSI were we
have also included the CDHS and CHARM data.
Global with NuTeV reanal.
NNPDF [163]
Bentz at al. [164]
NNPDF/Bentz et al.
NSI with down
NSI with up
NU
NU
−0.042 < εdV
µµ < 0.042
−0.044 < εuV
µµ < 0.044
−0.091 < εdA
µµ < 0.091
−0.15 < εuA
µµ < 0.18
−0.042 < εdV
µµ < 0.042
−0.044 < εuV
µµ < 0.044
−0.072 < εdA
µµ < 0.057
−0.094 < εuA
µµ < 0.140
FC
FC
−0.007 <
−0.039 <
εdV
µτ
εdA
µτ
< 0.007
−0.007 < εuV
µτ < 0.007
< 0.039
−0.039 < εuA
µτ < 0.039
three extra NSI couplings: εee , εeµ , and εeτ , making the analysis more complex and the
constraints weaker.
4.5
Summary
The original results reported by the NuTeV Collaboration suggested important deviations from the SM predictions. Considering that this could be a consequence of new
physics effects, we have reanalyzed the constraints on non-standard neutrino interactions
of muon-neutrinos with quarks. In particular, we have reanalyzed the results of the NuTeV
experiment introducing systematic uncertainties which were not taken into account in the
original analysis.
We have combined the restrictions obtained in this reanalysis with those coming from
atmospheric data, that were useful to remove degeneracies in the NSI parameter space.
Although we have found that muon-neutrino constraints are stronger than those for tau or
electron-neutrinos, they are weaker than previously believed. While bounds of the order
of few 10−3 are reported in the literature, when considering new uncertainties we found
constraints of the order of few 10−2 . These results are summarized in Table 4.2. For the
case of a general three-generation analysis, it is expected that even weaker constraints
will be found.
It is not expected that data from the MINOS long baseline experiment [218] could
114
Chapter 4. Probing νµ NSI with accelerator and atmospheric data
improve these results, since it has a reduced sensitivity to matter effects, especially when
compared to the atmospheric neutrino data. Since the same would be expected for the
OPERA experiment [219], we would need to wait for NOVA results [220] or the proposed
DUNE experiment [221] for an improvement in this sector.
Another natural direction would be the NuSOnG proposal [222], a new high statistics
neutrino scattering experiment [223], with the same goals as NuTeV. The experiment
would be devoted to probe muon-neutrino couplings and the expectation would be that
NuSOng would have twice the sensitivity of NuTeV. It is expected that NSI constraints
would be improved by about the same amount.
Chapter 5
Non-unitary lepton mixing matrix
As we saw, the neutrino oscillation phenomenon cannot be explained without nonzero
neutrino masses. The observed small neutrino mass differences can arise from an effective
dimension-five operator O5 ∝ LΦLΦ [16] (see Fig. 1.3) that violates the lepton number,
which may stem from unknown physics beyond the Standard Model. In the operator,
L denotes a lepton doublet while Φ is the SM scalar doublet. This operator generates
Majorana neutrino masses after electroweak symmetry breaking. In this the case, neutrino
mass appears in second order of hΦi, and lepton number violation by two units (∆L = 2)
can occur at large scale. This mechanism explains the small neutrino mass in comparison
to the Standard Model charged fermion masses 1 . However, the mechanism that generates
the operator O5 is still unknown and we cannot say much more about the interaction that
produces the operator or its associated mass scale, nor the possible details of its flavor
structure.
If there exist heavy “messenger” particles, the operator O5 could be induced through
the exchange of them. They could be neutral heavy leptons (NHL), that arise naturally
in several extensions of the Standard Model when Yν ∼ 1, playing a role as messengers of
neutrino mass generation. This is one of their strongest motivations and a key ingredient
of the Type-I seesaw mechanism [3] and variants. If this mechanism is realized at the
Fermi scale [18, 19, 22, 224–228], the “seesaw messengers” might give rise to several
interesting phenomenological implications, depending on the assumed gauge structure. If
we consider the minimal SU (3)C × SU (2)L × U (1)Y structure, we could have several new
states and potential signatures, for instance:
• Heavy isosinglet leptons. Below the Z mass, these particles have been searched
1
Notice that it has been already discussed in Sec. 1.4. Here, we return to it in order to introduce this
chapter properly.
115
116
Chapter 5. Non-unitary lepton mixing matrix
for at LEP I [229–231]; if their mass is in the TeV range, they might be seen in
the LHC experiments, although their rates are not expected to be large in the
SU (3)C × SU (2)L × U (1)Y gauge structure.
• Light isosinglet leptons, usually called “sterile” neutrinos. If the mass of these particles is found in the eV range they could help to explain current neutrino oscillation
anomalies [232, 233] since they will take part in the oscillations. If their mass is in
the keV range or above, they might be relevant for cosmology [234] or appear as
distortions in weak decay spectra [235].
• Universality violation. Due to the existence of NHL, weak decay rates would decrease if the NHL is not heavy enough to be emitted [236].
• Non-unitary lepton mixing matrix. Whenever NHL are too heavy to take part in
neutrino oscillations, the admixture with light neutrino states will affect the current
neutrino oscillation picture; in particular, the leptonic mixing matrix will be nonunitary [21].
• Neutrinoless double beta decay. If the NHL is a Majorana particle, lepton number
violation processes would appear, including the well-known neutrinoless double beta
decay. It could be generated through long-range (mass mechanism), or induce shortrange contributions [237–239].
• Charged lepton flavor violation processes. They could be induced by NHL, although
the rates and current constraints are model dependent [21, 240].
In this chapter we will consider the phenomenological consequences of a non-unitary
lepton mixing matrix. We will consider that the non-unitarity is produced by the existence
of NHL, for instance, from isosinglet neutrinos above 100 GeV. Therefore, these heavy
particles will be too heavy to take part in oscillations or low-energy weak decay processes.
We will show that the most general form of the mixing matrix is factorable, separating the
mixing matrix in a “new physics part” and a “standard part”. With this description, the
current experiments described with a two-neutrino scheme (electron and muon-neutrinos
for instance) can be expressed in terms of just four new parameters, three real mixing
angles plus one new CP violation phase. We will show in this chapter how these new
parameters affect oscillation probabilities, and we will discuss the main restrictions to
the mixing structure that follows from oscillation data as well as from universality tests.
We also present a compilation of several model-independent constraints on NHL mixing
5.1. A new way to describe extra heavy leptons
117
parameters, translated into this new parametrization. We consider in this last case both
low-energy searches and searches of direct NHL production at high-energy accelerator
experiments.
5.1
A new way to describe extra heavy leptons
The existence of isosinglet neutral heavy leptons implies that they could mix with the
standard isodoublet neutrinos. The mixing matrix including extra heavy leptons has
been studied carefully and described in the symmetric parametrization in Refs. [9, 45].
In this thesis, we consider an equivalent presentation that manifestly factorizes out the
parameters related with the extra heavy leptons and separate them from those describing
the light neutrino mixing within the unitarity approximation. We will present here the
main features of this parametrization and the more mathematical details will be shown
in the appendix B.
If we consider three light neutrinos (the conventional number of neutrinos in the SM)
and (n − 3) extra NHL, we have an U n×n unitary mixing matrix. We can divide this
mixing matrix in four different blocks [241]
U n×n =
N S
V T
!
,
(5.1)
where N is a 3 × 3 non-unitary matrix describing the light neutrino sector. The matrix S
describes the coupling of the isosinglet states, expected to be heavy 2 . We find that the
3 × 3 mixing matrix N can be decomposed most conveniently as a product of a standard
matrix U 3×3 and a “new physics matrix” N N P :

N = N N P U 3×3

α11 0
0


=  α21 α22 0  U 3×3 ,
α31 α32 α33
(5.2)
where U 3×3 is the usual unitary 3 × 3 leptonic mixing matrix reported in neutrino oscillation searches. The submatrix U 3×3 may be expressed in the symmetric parametrization
(Eq. (1.120)) or, equivalently, as prescribed by the Particle Data Group [6] 3 (Eq. (1.121)).
In this thesis we prefer to use the symmetric parametrization. The reader can find a discussion on the advantages of each parametrization in Ref [45]. On the other hand, the
2
3
See Ref. [17] for a perturbative expansion of U n×n .
The factorization holds irrespective of which form is taken for light neutrino mixing matrix.
118
Chapter 5. Non-unitary lepton mixing matrix
N N P matrix in Eq. (5.2) contains all the new physics information through the αij parameters, which are complex in general. As we can see, this N N P matrix induces the unitarity
violation. More explicitly, we can calculate the mixing matrix N obtaining:


α11 Ue1
α11 Ue2
α11 Ue3


α21 Ue1 + α22 Uµ1
α21 Ue2 + α22 Uµ2
α21 Ue3 + α22 Uµ3

.
α31 Ue1 + α32 Uµ1 + α33 Uτ 1 α31 Ue2 + α32 Uµ2 + α33 Uτ 2 α31 Ue3 + α32 Uµ3 + α33 Uτ 3
(5.3)
Notice the convenience of using Eq. (5.2), since it gives a general and complete description
of the current neutrino experiments when the unitarity condition is relaxed. An example
will show the use of this notation. If we consider interactions involving electron-neutrinos
and muon-neutrinos, we just need, with our description, four extra parameters which
include all additional information beyond the SM: the two real parameters α11 and α22 plus
a complex one, α21 , containing a single CP phase. For a discussion on the existence [242]
and possible effects [243] of extra CP phases associated to the admixture of NHL in
the charged leptonic weak interaction, the reader is referred to the original paper in [9].
The new point here is that, despite the proliferation of phase parameters, one can always
stack them in a single physical parameter per non-diagonal entry, so only one combination
enters in the “relevant” neutrino oscillation experiments. This interesting feature holds
independently of the number of extra NHL.
We write below the explicit form of the elements of the new physics matrix, αij , for
any number of extra NHL. They can be separated in:
• Diagonal elements, αii , which are real and expressed in a simple way as
α11
=
c1 n c 1n−1 c1 n−2 . . . c14 ,
α22
=
c2 n c 2n−1 c2 n−2 . . . c24 ,
α33
=
c3 n c 3n−1 c3 n−2 . . . c34 .
(5.4)
Notice that these diagonal terms depend only on the cosines of the mixing parameters, cij = cos θij .
• Off diagonal term α31 . In this case the expression is more complicated; however, if
we neglect quartic terms in sin θij , with j = 4, 5, · · · , we find:
α31 = c3 n c 3n−1 . . . c3 5 η34 c2 4 η̄14 + c3 n . . . c3 6 η35 c2 5 η̄15 c14 + . . .
+ η3n c2 n η̄1n c1n−1 c1n−2 . . . c14 ,
(5.5)
5.2. NHL and non-unitary neutrino mixing
119
where ηij = e−iφij sin θij and its conjugate, η̄ij = −eiφij sin θij , contain all of the CP
violating phases.
• Off-diagonal terms α21 and α32 , expressed as a sum of n − 3 terms as
α21 = c2 n c 2n−1 . . . c2 5 η24 η̄14 + c2 n . . . c2 6 η25 η̄15 c14 + . . .
+ η2n η̄1n c1n−1 c1n−2 . . . c14 ,
α32 = c3 n c 3n−1 . . . c3 5 η34 η̄24 + c3 n . . . c3 6 η35 η̄25 c24 + . . .
+ η3n η̄2n c2n−1 c2n−2 . . . c24 .
(5.6)
In summary, by choosing a convenient ordering for the products of the complex rotation
matrices ωij , one obtains a convenient parametrization that separates all the information
relative to the additional leptons in a simple and compact form, with a matrix containing
three zeroes and reducing the number of parameters related to new physics. In what
follows, we will show the utility of this specific parametrization and we will also obtain
the corresponding constraints.
5.2
NHL and non-unitary neutrino mixing
Keeping in mind the formalism described in the previous section and the chiral nature of
the SU (3)C × SU (2)L × U (1)Y model, we describe the couplings of the n neutrino states
in the charged-current weak interaction as part of a rectangular matrix K [9],
K=
N S
,
(5.7)
where the N block is the 3 × 3 matrix already discussed in Eq. (5.2), corresponding to
the three light neutrinos matrix with contributions of new physics, and the S block is a
3 × (n − 3) matrix corresponding to the mixing of extra heavy leptons.
In a scenario with extra heavy fermions that mix with the active light neutrinos, the
unitary mixing matrix is the complete matrix described by Eq. (5.1), so that the mixing
matrix involving the standard neutrinos comes from a truncation of Eq. (5.1). Therefore,
the 3 × 3 light neutrino mixing matrix N is not unitary and this fact modifies the SM
observables. Note also that, in this case, the unitarity condition takes the form
KK † = N N † + SS † = I ,
(5.8)
120
Chapter 5. Non-unitary lepton mixing matrix
with

2
α11

N N † =  α11 α21
α11 α31
∗
α11 α21
2
α22
+ |α21 |2
∗
α22 α32 + α31 α21

∗
α11 α31

∗
∗
α22 α32
+ α21 α31
.
2
α33
+ |α31 |2 + |α32 |2
(5.9)
The parametrization shown here allows to concentrate all the information of the (n−3)
heavy states into the αij parameters, obtaining a more compact and simple notation. The
parameters are totally general, include all the relevant CP phases and the parametrization
is model independent. In the following sections, we will show the advantages of this
parametrization by analyzing direct and indirect searches of the extra heavy leptons. We
will rewrite the relevant observables and will obtain the corresponding constraints on the
αij parameters.
5.3
Constraints from universality tests
Within the picture described in this chapter, the new extra heavy leptons will not participate in several weak processes 4 , since they can not be emitted due to kinematical
restrictions. As a consequence, the universality of certain weak processes will be broken
and the effective value of the Fermi constant will change accordingly; this is the case of
muon and beta decays, for example. Although the extra heavy leptons will not be kinematically emitted in various of these weak processes, they remain in the process as part of
the admixture with the light particles. So we can use the formalism characterized by the
αij parameters (Secs. 5.1 and 5.2) in order to describe various weak processes, translating their parameters in this compact formalism to derive the corresponding experimental
sensitivities.
First we will discuss the universality constraints. Let us start considering the current
bounds [236, 244–251], then we introduce them within the above formalism in order to
prove its simplicity. Comparing muon and beta decays,
A(Z, N ) → A(Z + 1, N − 1) + e− + ν̄e ,
A(Z, N ) → A(Z − 1, N + 1) + e+ + νe ,
(5.10)
µ+ → e+ + ν̄µ + νe ,
µ− → e− + νµ + ν̄e ,
4
Notice that neutral heavy leptons could participate virtually in weak processes.
(5.11)
5.3. Constraints from universality tests
121
for a model with just one extra heavy lepton, one finds the following corrections to the
effective Fermi constant [244, 245, 249],
Gµ = GF
q
(1 − |Se4
|2 )(1
− |Sµ4
|2 )
q
2
2
+ |α21 |2 )
(α22
= GF α11
(5.12)
and
Gβ = GF
p
(1 − |Se4 |2 ) .
(5.13)
Relaxing this assumption and considering the presence of n extra heavy leptons, we can
rewrite Eqs. (5.12) and (5.13) in terms of αij obtaining similar expressions,
q
p
2
2
†
†
+ |α21 |2 )
(α22
Gµ = GF (N N )11 (N N )22 = GF α11
(5.14)
q
p
2
†
Gβ = GF (N N )11 = GF α11
.
(5.15)
and
Notice that this change will affect all the observables related with the Fermi constant, for
example the quark CKM matrix elements [244]. We focus on the two matrix elements
Vud and Vus . Their new expressions are obtained from the vector coupling in β, Ke3 and
hyperon decays, after scaling by the new constant Gµ . The unitarity relation for the first
row of the CKM is now expressed as [244, 245]:
3
X
|Vui |2 =
i=1
Gβ
Gµ
2
=
!2
p
GF (N N † )11
1
p
=
,
†
†
(N N † )22
GF (N N )11 (N N )22
(5.16)
From this equation one gets the following constraint [6]:
3
X
i=1
|Vui |2 =
2
α22
1
= 0.9999 ± 0.0006 ,
+ |α21 |2
(5.17)
2
which implies that 1 − (N N † )22 = (SS † )22 = 1 − α22
− |α21 |2 < 0.0005 at 1σ.
We can also get constraints on these αij parameters from other universality tests. In
the absence of NHL the gauge bosons couple to the leptons with a flavor independent
strength. But in the presence of heavy isosinglets this is no longer true, and the couplings
will become flavor dependent because the unitarity of the mixing matrix has been broken
and N N † 6= I. We can use the reported ratios for these couplings to obtain constraints
from universality. As in the case of the CKM matrix elements, we may express them for
122
Chapter 5. Non-unitary lepton mixing matrix
Figure 5.1: Constraints from universality test for the unitarity deviations.
one extra heavy lepton as [249]
ga
gµ
2
=
(1 − |Sa4 |)
(1 − |Sµ4 |)
a = e, τ .
(5.18)
For a more general case they can be expressed as [244]
ga
gµ
2
=
(N N † )aa
(N N † )22
a = 1, 3 .
(5.19)
We can consider the pion decay branching ratio [247]:
Rπ =
Γ(π + → e+ ν)
.
Γ(π + → µ+ ν)
(5.20)
Notice that this ratio will be proportional to the previous one, shown in Eq. (5.19), for
the case a = 1 ≡ e. Then the current experimental value for the pion decay allows to
obtain the constraint [247, 252]:
2
Rπ
(N N † )11
α11
=
=
2
RπSM
(N N † )22
α22
+ |α21 |2
(1.2300 ± 0.0040) × 10−4
=
= 0.9956 ± 0.0040 .
(1.2354 ± 0.0002) × 10−4
rπ =
(5.21)
5.4. Neutrino oscillations with non-unitarity
123
2
2
+ |α21 |2 = 1. This
< 0.0084 at 1σ for the particular case of α22
This implies that 1 − α11
procedure was adopted in Ref. [249]. In general, [(SS † )22 ] 6= 0.
We can obtain a more robust constraint if we consider the unitarity constraint on the
CKM matrix already discussed. By combining the constraints from Eqs. (5.17) and (5.21)
we get the bounds shown in Fig. 5.1. This result can be translated into the constraints
2
< 0.0130 ,
1 − α11
2
− |α21 |2 < 0.0012 ,
1 − α22
(5.22)
at 90% C.L. for 2 d.o.f. We can notice that another measurement is needed to break
the degeneracy between the parameters α22 and |α21 |. In the next section, we will study
possible neutrino oscillation observables that could be of help for this analysis.
In order to complete this analysis, we can also find constraints from the µ − τ universality, which give information on the τ mixing parameters. Using Eq. (5.19) and the
corresponding experimental value [253], we obtain
(N N † )33
= 0.9850 ± 0.0057 .
(N N † )22
(5.23)
From this result it is straightforward to notice that 1 − (N N † )33 = (SS † )33 < 0.0207 at
1σ for the least conservative case of (SS † )22 = 0.
As we have seen, the existence of extra heavy leptons implies variations in all observables related to the Fermi constant. We have expressed these corrections in terms of a
few parameters that concentrate all the new physics information. In what follows we will
apply this formalism to the important phenomenon of neutrino oscillations.
5.4
Neutrino oscillations with non-unitarity
The compact notation with αij may be useful to study new physics effects in neutrino
oscillations. In this section we compute the expressions for this important case. Firstly
we will obtain new formulae for the neutrino survival and conversion probabilities and
then we will get restrictions for the new physics in terms of the α parameters. It should
be noted that the general expressions have been calculated neglecting cubic products of
α21 , sin θ13 and sin ∆21 , where ∆ji stands for ∆m2ji L/4E. This approximation is in a good
agreement with the current experimental status.
124
Chapter 5. Non-unitary lepton mixing matrix
First of all, we recall the most general expression for the neutrino oscillation probabilities in vacuum [12] (Sec. 1.5.1),
Pνα →νβ =
3
X
∗
∗
Uαi
Uβi Uαj Uβj
i,j
3
X
∆m2ji L
− 4
Re
sin
4E
j>i
3
X
∗
∆m2ji L
∗
+ 2
Im Uαj Uβj Uαi Uβi sin
,
2E
j>i
∗
∗
Uαj
Uβj Uαi Uβi
2
(5.24)
where, if we introduce the unitarity condition of the mixing matrix, U U † = I, we obtain
the usual expression in the literature,
Pνα →νβ = δαβ
3
X
∆m2ji L
− 4
Re
sin
4E
j>i
3
X
∗
∆m2ji L
∗
.
+ 2
Im Uαj Uβj Uαi Uβi sin
2E
j>i
∗
∗
Uαj
Uβj Uαi Uβi
2
(5.25)
For the case with extra neutral heavy leptons, unitarity is broken, N N † 6= I 5 . Although NHL will not participate in the oscillation because they are too heavy to be
produced, they will mix with light neutrinos, so we have a similar expression as Eq. (5.24)
but using N as the mixing matrix,
Pνα →νβ =
3
X
i,j
∗
∗
Nαi
Nβi Nαj Nβj
3
X
∆m2ji L
− 4
4E
j>i
3
X
∗
∆m2ji L
∗
+ 2
Im Nαj Nβj Nαi Nβi sin
.
2E
j>i
∗
2
∗
Re Nαj
Nβj Nαi Nβi
sin
(5.26)
If we consider the muon-neutrino conversion into an electron-neutrino, the probability
is:
5
Notice again that N is the mixing matrix following from the truncation of the general matrix U ,
Eq. (5.1).
5.4. Neutrino oscillations with non-unitarity
Pνµ →νe =
3
X
∗
∗
Nµi
Nei Nµj Nej
i,j
3
X
125
∆m2ji L
− 4
Re
sin
4E
j>i
3
X
∗
∆m2ji L
∗
.
+ 2
Im Nµj Nej Nµi Nei sin
2E
j>i
∗
Nµj
Nej Nµi Nei∗
2
(5.27)
Using the condition given in Eqs. (5.8) and (5.9) for the 3 × 3 matrix N , instead of
the usual unitarity condition, we observe how the unitarity is broken. The Dirac delta
disappears and instead of it we have the new αij terms:
Pνµ →νe =
2
α11
|α21 |2
3
X
∆m2ji L
− 4
4E
j>i
3
X
∗
∆m2ji L
∗
+ 2
Im Nµj Nej Nµi Nei sin
.
2E
j>i
∗
Re Nµj
Nej Nµi Nei∗ sin2
(5.28)
In order to obtain a more explicit expression of Eq. (5.28), we may use Eq. (5.3) and
substitute the values of Nαi in terms of Uαi and αij , arriving to the expression:
Pνµ →νe
!
2
L
∆m
ji
2
= α11
|α21 |2 1 − 4
|Uej |2 |Uei |2 sin2
4E
j>i
3
X
∗
2 ∆m2ji L
2
∗
− (α11 α22 ) 4
Re Uµj Uej Uµi Uei sin
4E
j>i
3
X
∗
∆m2ji L
2
∗
+ (α11 α22 ) 2
Im Uµj Uej Uµi Uei sin
2E
j>i
3
X
3
X
∆m2ji L
−
4E
j>i
3
X
∆m2ji L
2
2 ∗
∗
2
∗
.
+ 2α11 α22
Im α21 |Uei | Uµj Uej + α21 |Uej | Uµi Uei sin
2E
j>i
2
4α11
α22
∗
∗
Re α21 |Uei |2 Uµj
Uej + α21
|Uej |2 Uµi Uei∗ sin2
(5.29)
We can now use the standard parametrization for the mixing matrix entries (Uαi ) in this
equation. We remind the reader that we neglect cubic products of α21 , sin θ13 , and sin ∆21 .
126
Chapter 5. Non-unitary lepton mixing matrix
With this approximation one obtains
2
2
Pνµ →νe = (α11 α22 )2 Pν3×3
+ α11
α22 |α21 |PνIµ →νe + α11
|α21 |2 ,
µ →νe
(5.30)
denotes the standard neutrino conversion probability for the three active
where Pν3×3
µ →νe
species [189, 254, 255],
Pν3×3
µ →νe
∆m221 L
4 cos θ12 cos θ23 sin θ12 sin
4E
2
∆m31 L
2
2
2
2
cos θ13 sin θ13 sin θ23 sin
4E
∆m221 L
sin(2θ12 ) sin θ13 sin(2θ23 ) sin
2E
2
2
∆m31 L
∆m31 L
sin
− I123 ,
cos
4E
4E
=
+
+
×
2
2
2
2
(5.31)
with I123 = −δCP = φ12 − φ13 + φ23 . On the other hand, PνIµ →νe denotes a new physics
term which depends on an extra CP phase, INP = φ12 − Arg(α21 ), along with the usual
mixing angles and the standard CP phase through I123 ,
PνIµ →νe
∆m231 L
∆m231 L
= −2 sin(2θ13 ) sin θ23 sin
+ IN P − I123
sin
4E
4E
∆m221 L
− cos θ13 cos θ23 sin(2θ12 ) sin
sin(IN P ) .
(5.32)
2E
We would like to remark again that the neutrino conversion probability in this channel
depends on two CP phases. Besides the standard CP phase, there is a new phase that
comes from the existence of the NHL, denoted as IN P . This “new physics” phase condensates (for this channel) the impact of all the additional phases from NHL and appears in
our formalism as the phase of the parameter α21 . IN P appears in both terms of Eq. (5.32),
in the first one as a difference with the standard phase, I123 − IN P , being proportional
to sin θ13 . In the second one, IN P is not accompanied by the standard phase and the
term depends on the solar mass difference ∆m221 . Both terms are expected to give small
corrections to the standard probability.
The contribution of the extra neutral heavy lepton to the neutrino oscillation data can
be shown qualitatively by giving specific values to the new parameters introduced here.
For this purpose, we consider the case where α11 = 1, α22 = 0.9939, |α21 | = 0.078, and IN P
is equal to π/2, 3π/2, π or 0. We have chosen the standard phase δ = −I123 = 3π/2, close
5.4. Neutrino oscillations with non-unitarity
127
to its best fit [59]. With these values, the conversion probability takes the form shown
in Fig. 5.2 which shows that the new CP phase could be interesting in future neutrino
experiments.
0.2
Standard
INP = 3π/2
INP = π/2
P(νµ −> νe)
0.15
Standard
INP = π
INP = 0
0.1
0.05
0
100
500
1000 2000
L[km]
500
1000 2000
L[km]
Figure 5.2: Neutrino conversion probability for Eν = 1 GeV. We show in this plot the
standard conversion probability (black) for the CP phase δ = −I123 = 3π/2. The nonunitary cases are illustrated for the particular values α11 = 1, α22 = 0.9939, and |α21 | =
0.078. In the left panel we consider the following values for the new CP phase, IN P = π/2
(magenta line) and IN P = 3π/2 (green line). The cases IN P = 0 (magenta line) and
IN P = π (green line) are shown in the right panel.
For the muon neutrino survival probability, we proceed in a similar way as we did in
the Pνµ →νe case, getting the expression
Pνµ →νµ =
3
X
i,j
∗
∗
Nµi
Nµi Nµj Nµj
−4
3
X
j>i
∗
2
∗
Re Nµj
Nµj Nµi Nµi
sin
∆m2ji L
4E
.
(5.33)
Substituting again the values of Nαi from Eq. (5.3), we can appreciate once more a decrease
128
Chapter 5. Non-unitary lepton mixing matrix
in the probability before neutrinos have a chance to oscillate, the zero distance effect [83]:
2
Pνµ →νµ = (|α21 | +
2 2
α22
)
−4
3
X
2
2
2
|Nµj | |Nµi | sin
j>i
∆m2ji
L ,
4E
(5.34)
2 2
)
Pνµ →νµ = (|α21 |2 + α22
3
X
∆m2ji
2
2
2
|α21 Uej + α22 Uµj | |α21 Uei + α22 Uµi | sin
L .
− 4
4E
j>i
(5.35)
As we did previously, we neglect for simplicity cubic products of α21 , sin θ13 , and sin ∆21 ,
obtaining the following expression:
4
3
2
Pν3×3
+ α22
|α21 |PνIµ1→νµ + 2|α21 |2 α22
Pνµ →νµ = α22
PνIµ2→νµ ,
µ →νµ
(5.36)
where Pν3×3
corresponds to the standard oscillation formula,
µ →νµ
Pν3×3
µ →νµ
≈
+
×
−
+
2
∆m231 L
2
2
2
2
1 − 4 cos θ23 sin θ23 − cos(2θ23 ) sin θ23 sin θ13 sin
4E
2
2
2
3
2 cos θ12 cos θ23 sin θ23 − cos(I123 ) cos θ23 sin(2θ12 ) sin θ23 sin θ13
∆m221 L
∆m231 L
sin
sin
2E
2E
∆m231 L
2
2
2
4 cos θ12 cos θ23 sin θ23 cos
2E
∆m221 L
2
4
2
2
cos θ12 cos θ23 sin θ12 sin
.
(5.37)
4E
The two extra contributions to the standard probability are given by:
PνIµ1→νµ
PνIµ2→νµ
∆m231 L
≈ − 8 [sin θ13 sin θ23 cos(2θ23 ) cos(I123 − INP )] sin
4E
2
∆m31 L
∆m221 L
2
+ 2 cos θ23 sin(2θ12 ) sin θ23 cos(INP ) sin
sin
,
2E
2E
∆m231 L
2
2
≈ 1 − 2 sin θ23 sin
.
(5.38)
4E
2
We show graphically this probability in Fig. 5.3.
Since the contribution in this case is very small and, therefore, difficult to appreciate,
we have decided to plot the difference between the probability including new physics and
5.4. Neutrino oscillations with non-unitarity
129
1
Standard
INP = 3π/2
INP = π/2
P(νµ −> νµ)
0.8
Standard
INP = π
INP = 0
0.6
0.4
0.2
0
100
500
1000 2000
500
L[km]
1000 2000
L[km]
Figure 5.3: Survival probability for a fixed neutrino energy of Eν = 1 GeV. We show the
standard survival probability, with δ = −I123 = 3π/2 in the black curve. The non-unitary
case is also given for the same parameter values of Fig. 5.2.
P
the standard probability (PνNµ →ν
− Pν3×3
) in Fig. 5.4.
µ
µ →νµ
One can see from these plots that the disappearance channel has worse sensitivity
to the extra CP phase IN P than the appearance channel. The computations in these
plots were done for the same values of the appearance case: α11 = 1, α22 = 0.9939,
|α21 | = 0.078, δ = −I123 = 3π/2, and IN P equal to either π/2 or 3π/2.
We turn now to terrestrial experiments using reactors or radioactive sources, and
discuss oscillations of electron-neutrinos or antineutrinos. They are relevant in the description of solar neutrino experiments. The electron-(anti)neutrino survival probability
in vacuum is given by the following expression:
Pνe →νe =
3
X
∗
Nei∗ Nei Nej Nej
i,j
−4
3
X
Re
∗
Nej
Nej Nei Nei∗
sin
2
j>i
∆m2ji L
4E
.
(5.39)
For this case, it is easy to notice that Nei = α11 Uei (see for example Eq. (5.3)) and,
therefore,
"
4
Pνe →νe = α11
3
X
i
|Uei |2 |Uei |2 − 4
3
X
j>i
|Uej |2 |Uei |2 sin2
#
∆m2ji
L .
4E
(5.40)
Chapter 5. Non-unitary lepton mixing matrix
INP = 0
INP = π
INP = π/2
INP = 3π/2
10
0
NP
3x3
(Pµµ - Pµµ ) × 10
3
130
-10
100
200
1000
500
2000
L[km]
200
500
1000
2000
L[km]
Figure 5.4: Difference between the standard muon-neutrino survival probability and the
corresponding new physics probability for some values of the additional CP phase IN P .
The remaining parameters have been fixed to be the same as in Fig. 5.2.
Using Eq. (1.120), this expression can be rewritten as
Pνe →νe =
4
α11
4
2
2
1 − cos θ13 sin (2θ12 ) sin
∆m231
∆m221
2
2
L − sin (2θ13 ) sin
L .
4E
4E
(5.41)
Notice that, for this last probability, the existence of neutral heavy leptons produces
4
a correction given simply by an overall factor α11
. This parameter will contain the effects
of unitarity violation. Unfortunately, due to the strong restrictions on universality, the
expected modifications to the probability will be limited.
To conclude this section, we point out that this formalism can easily be applied to
the case of extra light neutrinos. They will take part in oscillations and may play a
role [256, 257] in the well-known anomalies reported by reactor neutrino experiments [258]
and the MiniBooNE Collaboration [233].
5.5
Neutrino oscillations and universality
As a consequence of the effective non-unitarity of the 3 × 3 leptonic mixing matrix, one
may observe neutrino flavor change even though neutrinos have not traveled yet [83],
a phenomenon known in the literature as the zero distance effect. The survival and
conversion probabilities differ from one and zero respectively, as can be seen from Eqs.
5.5. Neutrino oscillations and universality
131
(5.29), (5.34) and (5.40). In those equations, instead of the usual Dirac delta (δαβ ) we
find α terms, which are independent of the distance (L). Therefore, for zero distance we
express these probabilities in terms of αij as
Pνe →νe = [α11 ]4 = [(N N † )11 ]2 = [1 − (SS † )11 ]2 ,
2 2
] = [(N N † )22 ]2 = [1 − (SS † )22 ]2 ,
Pνµ →νµ = [|α21 |2 + α22
(5.42)
2
Pνµ →νe = α11
|α21 |2 = [(N N † )21 ]2 = [(SS † )21 ]2 .
From these expressions we can make an estimate of the constraints for the new parameters. We first notice that it is possible to express these equations in a similar way to the
case of a light sterile neutrino in the limit ∆m2ij L/(4E) 1 ( sin2 (∆m2ij l/(4E)) = 1/2).
Analogous expressions are obtained [259]:
1 2
sin (2θee ) eff ,
2
1
= 1 − sin2 (2θµµ ) eff ,
2
1 2
=
sin (2θµe ) eff ,
2
Pνe →νe = 1 −
Pνµ →νµ
Pνµ →νe
(5.43)
with
2
4
sin (2θee ) eff = 2(1 − α11
),
2
2 2
sin (2θµµ ) eff = 2[1 − (|α21 |2 + α22
) ],
2
2
sin (2θµe ) eff = 2α11
|α21 |2 .
(5.44)
Although the MiniBooNE experiment reports an anomaly that might be interpreted as
new physics [233], there are strong constraints on light sterile neutrinos at 3σ [257] given
by:
sin2 (2θee )
eff
≤ 0.2 ,
sin2 (2θµµ ) eff ≤ 0.06 ,
2
sin (2θµe ) eff ≤ 1 × 10−3 .
(5.45)
132
Chapter 5. Non-unitary lepton mixing matrix
Even stronger constraints on the existence of a fourth neutrino are derived from νµ to νe
oscillation experiments such as NOMAD [260]:
2
sin (2θµe ) eff ≤ 1.4 × 10−3 .
(5.46)
Translated into the parametrization under discussion we have
2
|α21 |2 ≤ 0.007 (90% C.L.) .
α11
(5.47)
Finally, we can obtain bounds at 90 % C.L. on the individual αij parameters, by
combining this limit with those coming from universality (Eq. (5.22)):
2
α11
≥ 0.987,
5.6
2
α22
≥ 0.9918,
|α21 |2 ≤ 0.0071.
(5.48)
Current constraints on NHL
The unitarity violation discussed in this chapter could point to new physics as the one
responsible for neutrino mass. For example, the possible existence of neutral heavy leptons
could give an explanation of the neutrino mass generation mechanism in seesaw schemes.
They might be the messengers in this mechanism. However, in these schemes very heavy
right-handed neutrinos are predicted, restricting their expected signatures. Nevertheless,
several low-scale seesaw realizations do not introduce these high-scale masses [18, 19, 22,
224–226]. For this reason, we consider useful to give a compilation of model-independent
NHL searches. These results are based only on a phenomenological analysis without
taking into account any particular seesaw scheme.
For many years, experiments have been searching for isosinglet neutrinos without any
positive signal up to now. The range of masses already covered is wide, between 10 MeV
and 100 GeV, looking for:
• A peak in leptonic decays of pions and kaons. It could imply a mixing of heavy
neutrinos with electron or muon-neutrinos [249, 261]. The signal of a neutral heavy
lepton would be a monochromatic line at
El =
m2M + m2l − m2N HL
,
2mM
(5.49)
where El and ml are the energy and mass of the lepton respectively, mM is the mass
of the meson and mN HL corresponds to the mass of the heavy neutrino.
5.6. Current constraints on NHL
10
133
0
NA3
ATLAS
Belle
-2
10
LEP2
-4
|Se j|
2
10
L3
K --> e ν
DELPHI
-6
10
10
π −−> e ν
CHARM
-8
PS191
-1
10
10
0
10
1
10
2
10
3
mj (GeV)
Figure 5.5: Bounds on the mixing between the electron-neutrino and neutral heavy
leptons, |Sej |, depending on the NHL mass, mj .
• A signature coming from a possible heavy neutrino decay in electrons, muons and
pions.
In particular, for the couplings of heavy neutrinos with electron and muon-neutrinos,
the mixing parameters |Sej |2 and |Sµj |2 can be tested (between 10 MeV and 100 GeV) in:
• Peak searches in the energy spectrum of leptons (electron or muon) in meson decays:
– π → eν and K → eν for the coupling with electron-neutrino.
– π → µν and K → µν for the muon-neutrino case.
• NHL decays in reactor and accelerator neutrino experiments. If they were heavy
enough but lighter than the Z boson, they would have been copiously produced
in the first phase of the LEP experiment (the coupling should be appreciable [229,
230]). Searches have been negative, including those performed in the second phase
of LEP with higher energies [231] or in LHC [262–264].
We show the constraints coming from direct production of NHL in Fig. 5.5 and Fig. 5.6.
In Fig. 5.5 we show the wide range of energy that has been scanned in the search
for neutral heavy leptons coupled to the electron-neutrino. The constraints on |Sej |2
are given as a function of mj and can be interpreted as bounds on the m4 − Ue4 plane.
Figure 5.5 shows the results from nine different experiments [194, 231, 262, 265–273], since
134
Chapter 5. Non-unitary lepton mixing matrix
10
0
NA3
CHARM II
-2
10
-4
|Sµj|
ATLAS
LHCb
K -> µ ν
BELLE
2 10
DELPHI
-6
FMMF
10
10
CMS
L3
BEBC
NuTeV
-8
E949
0.1
PS191
1
10
mj (GeV)
100
Figure 5.6: Bounds on the mixing between the muon-neutrino and neutral heavy leptons,
|Sµj |, depending on the NHL mass, mj .
0
10
-1
10
-2
10
2 -3
|Sτj| 10
-4
10
-5
10
-6
10 -2
10
NOMAD
CHARM
DELPHI
-1
10
0
10
1
10
2
10
m j(GeV)
Figure 5.7: Bounds on the mixing between the tau-neutrino and neutral heavy leptons,
|Sτ j |, depending on the NHL mass, mj .
the CHARM Collaboration up to the recent LHC results from ATLAS. A brief description
of all these experiments is given in the appendix C. Future experiments will also search
for this heavy lepton [274]. That is the case, for instance, of DUNE [221], ILC [275] and
FCC-ee [276, 277] proposals.
5.6. Current constraints on NHL
135
We show in Fig. 5.6 the constraints for the mixing of extra neutral heavy leptons with
the muon-neutrino. The constraints also cover a wide region of energy as result of several
experimental searches [263, 264, 278–284]. Again, a short description of the experiments
that have constrained this coupling is given in the appendix C.
For the sake of completeness, we also show in Fig. 5.7 the restrictions for the neutral
heavy lepton mixing with the tau-neutrino [231, 285, 286]. By comparing with previous
figures, one can see that, as expected, the tau-neutrino sector has been less explored.
Colliders such as the LHC can set more stringent constraints thanks to the ATLAS and
CMS detectors. However, it should also be noticed that for the SU (3)C × SU (2)L × U (1)Y
gauge structure, extra heavy leptons will be isosinglets and, therefore, their production
rate will be scaled by small mixing factors. In order to avoid this limitation, one can
consider extended models, such as the left-right symmetric models (Sec. 1.4.4). In this
type of schemes extra right-handed gauge bosons could be produced at high energies,
opening the door to lepton flavour violation signatures [287, 288].
5.6.1
Neutrinoless double beta decay
It is expected, based on theoretical grounds, that neutrinos have a Majorana nature which
implies lepton number violation. In particular, neutrinoless double beta decay (0νββ) may
take place at some level [237].
We begin this section by recalling that the effective Majorana neutrino mass [289] is
given by
X
n×n 2
hmi = |
(Uej
) mj |,
(5.50)
j
where the j index indicates the light neutrinos couplings to the electron and the W -boson.
This dependence on the mixing matrix U n×n implies that the presence of the heavy
n×n
neutrinos would change the charged-current couplings to Uej
= α11 Uej . This is derived
from Eqs. (5.2) and (5.3), modifying Eq. (5.50) to
hmi = |
X
(α11 Uej )2 mj |,
(5.51)
j
with the index j running for the three light neutrino species.
The presence of an extra heavy neutrino in the neutrinoless double beta decay will
induce an extra amplitude contribution to this process, that will depend on the mixing
with the electron component. The amplitude will involve the exchange of the heavy
136
Chapter 5. Non-unitary lepton mixing matrix
Figure 5.8: Diagram of neutrinoless double beta decay. A virtual Majorana neutrino is
exchanged.
Majorana neutrinos Fig. 5.8 and it will be proportional to
A∝
q2
mj
,
− m2j
(5.52)
where q is the virtual momentum transfer (q ∼ 100 MeV). It is well known that there are
two different regimes for this amplitude [45]; if the virtual neutrino particle is light, we
can consider q 2 m2j and we have
Alight ∝ mj .
(5.53)
On the other hand, for the exchange of a heavier neutrino we have q 2 m2j , leading to
Aheavy ∝
1
.
mj
(5.54)
From current constraints on neutrinoless double beta decay it is possible to obtain a
restriction on the effect of an extra heavy lepton with a mass mj and a mixing Sej . Using,
for instance, the data from 76 Ge searches and assuming an isosinglet neutrino of mass
mj [290], it is possible to obtain constraints as shown in Fig. 5.9. In this figure we may see
the change of slope when the mass of the NHL approaches the nuclear momentum transfer,
mj ' 100 MeV, as expected from the previous discussion. Although the restrictions
5.6. Current constraints on NHL
137
0
10
-2
10
-4
|Se j|
2
10
-6
10
0νββ
-8
10
-1
10
0
10
1
10
2
10
mj (GeV)
Figure 5.9: Constraints from neutrinoless double beta decay to an isosinglet of mass mj ,
depending on its electron neutrino component |Sej |.
coming from this process are tighter than those shown in Figs. 5.5, 5.6 and 5.7, it should
be reminded that neutrinoless double beta decay takes place only if the neutrinos are
Majorana particles.
5.6.2
Charged lepton flavor violation
For the sake of completeness, we comment the charged lepton flavor violation processes.
These processes could be induced by the exchange of a virtual NHL, either at lowenergies [21] or at the high-energy scale of accelerator experiments [240]. They could
be seen at hadronic colliders, such as the LHC, especially if we consider well motivated
low-scale seesaw models [287, 288, 291, 292]. However, the rates of these processes would
depend on additional flavor parameters as well as upon details on the seesaw mechanism
providing masses to neutrinos, being very model-dependent. Since we have performed a
general model-independent analysis we have not considered charged lepton flavor violation
in this thesis.
138
5.7
Chapter 5. Non-unitary lepton mixing matrix
Summary
Extensions of the Standard Model, such as models employing the seesaw mechanism,
predict the breaking of weak universality and a non-trivial leptonic mixing matrix. This
can be generated by the admixture of heavy isosinglet neutrinos in leptonic chargedcurrents. These models are well motivated and most of them are successful in generating
neutrino mass and mixing, thanks to the presence of the NHL. These heavy particles acts
as messengers that generate the small neutrino masses. The scale of such heavy righthanded neutrinos is huge, but other realizations of these models exist, like the low-scale
seesaw models. These low-scale models suggest that such NHL may be light enough to be
accessible at high-energy colliders such as the LHC, or indirectly induce sizeable unitarity
deviations in the “effective” lepton mixing matrix.
When we study extended models where extra heavy neutrinos are included, we work
with many parameters, making it difficult to describe the model and their implications.
Using our parameterization (Eq. (5.2)) simplifies the description of such models, since it
involves only a set of effective parameters (three real plus a single CP phase) that concentrate all the new physics information. With this parameterization we have described the
impact of non-unitary lepton mixing on weak decay processes and neutrino oscillations.
Finally, for completeness, we have shown the up-to-date status of direct searches for extra
heavy leptons, translated into this notation.
Chapter 6
Conclusions
Particle physics is going through an interesting epoch, with important discoveries in the
past two decades. As a confirmation of this, we find the recent recognition to this field
through the Nobel prizes awarded to the Higgs boson discovery (2013) and the detection
of neutrino oscillations (2015). Particle physics is in an era of precision, prepared to solve
new challenges raised in our understanding of the universe.
At these crossroads of knowledge is situated a little and seemingly inoffensive particle,
the neutrino. It may seem insignificant but it has great news to tell us, new physics is
here! Neutrino oscillations are a solid evidence of physics beyond Standard Model. But it
is not an isolated phenomenon since it is accompanied by another prodigy, neutrinos are
massive particles! This implies the necessity to develop mechanisms and models to induce
neutrino mass. From most of these models we have non-conventional neutrino properties
like Non-Standard Interactions (NSI) or the magnetic moment of neutrino. Also processes
such as neutrinoless double beta decay, since typically neutrinos are Majorana. This thesis
has tried to derive information on NSI and their implications, using the experimental data
from oscillation searches.
For this purpose, we have started with two introductory chapters:
• In chapter 1 we have reviewed the most interesting features of neutrinos in the
Standard Model and beyond.
• In chapter 2 we have introduced the physics of Non-Standard Interactions, reviewing
how NSI theoretically arise and their effects in neutrino experiments.
After these introductory chapters we continue with the original part of this thesis
where the NSI effect on oscillation experiments has been discussed. In particular, a global
analysis of solar data including KamLAND reactor data has been performed in order to
139
140
Chapter 6. Conclusions
check how robust is the LMA solution to the solar neutrino problem in presence of NSI.
After introducing NSI, an extra solution appears, the so-called LMA-D (Large Mixing
Angle-Dark). It implies that solar neutrino data can also be accounted for θ12 > π/4
if the NSI strength is large enough. Therefore, although NSI play a sub-leading role in
the explanation of the solar neutrino problem, they still have influence over the neutrino
oscillation solution. Thus a question arises immediately: how is it possible to eliminate the
degeneracy of LMA? This a challenge for solar experiments like Borexino, SNO or SuperKamiokande and for new proposals as DUNE [221], LENA [293] or another LBNE (Long
Baseline Neutrino Experiments) [188, 294]. Atmospheric and laboratory experiments such
as Hyper-Kamiokande [206, 207] IceCube and DeepCore [295, 296] will be help in ruling
out the “dark” region.
Following with the study of NSI effects on neutrino experiments, a new analysis of
NuTeV results has been performed in this thesis. NuTeV results present a discrepancy
with the Standard Model parameters, taken as a possible hint of NSI. Since several uncertainties from QCD were omitted, we have carried out a reanalysis of NuTeV results in this
thesis. Several bounds on NSI couplings involving muon-neutrinos and quarks (εqµα ) have
been derived from our discussion. Our conclusion is that, although these NSI parameters
are much smaller than those involving electron and tau-neutrinos, they need not be as
small as previously believed.
These studies have allowed us to obtain limits on vectorial and axial components of
NSI parameters in processes of neutrinos with quarks (εqV,A
αα ). In particular, NSI couplings involving electron and tau-neutrinos have been studied by combining solar, reactor
(KamLAND) and accelerator (CHARM) data; whereas muon-neutrino NSI parameters
have been analyzed using the results from the NuTeV reanalysis, combining them with
accelerator (CHARM and CDHS) and atmospheric data (Super-Kamiokande) in order to
lift the degeneracy. The most relevant results are compiled in Table 6.1 at 90% of C.L.
In the last chapter of this thesis we have assumed the existence of neutral heavy leptons
and studied their implications. These heavy particles arise in most extended models as
exchange particles in the mechanism generating neutrino masses. As a consequence, NSI
could appear directly or indirectly through the non-unitarity of the light neutrino mixing
matrix, as we expressed in Eqs. (6.1). These models beyond the SM usually imply a
very large number of parameters, making it difficult to work with them. In order to
facilitate the study of these models, we have presented a model-independent formalism
which provides a very useful way to separate the light mixing matrix (N 3×3 ) in two parts:
the standard one (U SM ≡ U 3×3 ) and another one which concentrates the new physics
141
Table 6.1: Limits on NSI parameters of neutrino with quark interactions at 90 % C.L.
following from several combined analysis. The information on this table corresponds to
original results derived in this thesis.
NSI with down
NSI with up
NU
NU
Experiments
−0.5 < εdV
ee < 1.2
Sol+KamL+CHARM
−0.4 < εdA
ee < 1.4
Sol+KamL+CHARM
−1.8 <
−1.5 <
−0.042 <
−0.091 <
−0.042 <
−0.072 <
εdV
ττ
εdA
ττ
dV
εµµ
εdA
µµ
dV
εµµ
εdA
µµ
< 4.4
Sol+KamL+CHARM
< 0.7
Sol+KamL+CHARM
< 0.042 −0.044 < εuV
µµ < 0.044
< 0.091
−0.15 <
< 0.042 −0.044 <
< 0.057
FC
−0.094 <
εuA
µµ
uV
εµµ
εuA
µµ
NuTeV(NNPDF)+SK
< 0.18
NuTeV(NNPDF)+SK
< 0.044
NuTeV(Bentz et al.)+SK
< 0.140
NuTeV(Bentz et al.)+SK
FC
−0.08 < εdV
eτ < 0.58
−0.007 <
−0.039 <
εdV
µτ
εdV
µτ
Sol+KamL
< 0.007 −0.007 <
< 0.039
−0.039 <
εuV
µτ
εuA
µτ
< 0.007
NuTeV+SK+CHARM+CDHS
< 0.039
NuTeV+SK+CHARM+CDHS
information (N N P ). This formalism has been applied to:
• obtain bounds on the relevant αij parameters from universality tests,
• describe the oscillation probabilities, observing interesting features, such as zerodistance effects and the appearance of a new phase IN P which could affect the
oscillation probability,
• describe (3 + 1) and (3 + 3) neutrino schemes 1 .
These applications show explicitly the simplicity of this method. As a future project, it
would be very convenient to rewrite NSI couplings in terms of the αij parameters, for
instance:
|α21 ||α31 |
|α21 ||α31 |
, εeR
,
(6.1)
εeL
ee = −gL
ee = −gR
|α32 |α22
|α32 |α22
where a scenario with an extra heavy neutrino is considered.
1
See appendix B.
142
Chapter 6. Conclusions
Figure 6.1: Estimated LENA sensitivity to the non-unitary parameters [297].
For the sake of completeness, a compilation of the mixing parameters |Sαj |2 of heavy
neutrinos with the light ones from several laboratory experiments has been carried out.
We want to stress that studying these bounds could be very useful in order to elucidate
the neutrino mass generation mechanism and the energy scale of new physics. But it is
not possible from current experiments to get very strong bounds on these parameters,
so new experimental proposals must be considered, such as the aforementioned LENA.
As an example, an estimate of the LENA sensitivity to the non-unitarity of the mixing
matrix is shown in Fig. 6.1 in terms of α11 .
Summarizing in a couple of sentences: Neutrinos have opened the door to new physics,
involving non-standard effects and new extra particles among other features. Faced with
these prospects, we should be aware that a very interesting neutrino era is waiting for us.
Conclusions
La fı́sica de partı́cules es troba ara mateix en una època interessant, amb importants
descobriments durant les dues dècades passades. Com a prova d’açò, trobem el recent
reconeixement als premis Nobel, atorgant aquesta distinció al descobriment del bosó de
Higgs (2013) i a la detecció de les oscil·lacions de neutrins (2015). La fı́sica de partı́cules
es troba immersa en una era de precisió, preparada per resoldre els nous reptes plantejats
en el nostre intent d’entendre l’univers.
En meitat d’aquesta cruı̈lla de coneixements es troba una diminuta i aparentment
inofensiva partı́cula, el neutrı́. Sembla insignificant però té notı́cies importants per a
nosaltres, la nova fı́sica està ja acı́! Les oscil·lacions de neutrins són una sòlida evidència
de fı́sica més enllà del Model Estàndard. Però aquest no és un fenomen aı̈llat sinó que ve
acompanyat d’altre prodigi, els neutrins són partı́cules amb massa! Aquest fet implica la
necessitat de desenvolupar mecanismes i models que indueixin la massa del neutrı́. De la
gran majoria d’aquests models sorgeixen propietats no convencionals dels neutrins com
les Interaccions No Estàndard (NSI) o el moment magnètic. També processos com la
desintegració doble beta sense neutrins, ja que els neutrins són tı́picament partı́cules de
Majorana. Com el lector haurà comprovat, aquesta tesi ha intentat aportar informació
sobre les NSI i les seves implicacions, utilitzant les dades experimentals provinents de la
cerca d’oscil·lació de neutrins.
Per a aquest comès, hem començat la tesi amb dos capı́tols recopilatoris:
• El capı́tol 1 ha sigut una revisió de les propietats més importants dels neutrins dins
del Model Estàndard i més enllà d’aquest.
• El capı́tol 2 ens ha endinsat en la fı́sica de les Interaccions No Estàndard, revisant
el seu origen a la teoria i la seva influència als experiments de neutrins.
Després d’aquests dos capı́tols introductoris hem continuat amb la part original d’aquesta
tesi, on l’efecte de les NSI sobre experiments d’oscil·lació de neutrins ha sigut discutit.
En concret, s’ha realitzat una anàlisi global utilitzant les dades provinents d’experiments
143
144
Chapter 6. Conclusions
amb neutrins solars i incloent els resultats de l’experiment de reactor KamLAND, per
tal de dilucidar com de robusta és la solució LMA al problema dels neutrins solars en
presència de NSI. Després d’introduir les NSI apareix una solució extra no contemplada
anteriorment, la anomenada LMA-D (Large Mixing Angle-Dark). Aquesta nova solució
implica que les dades de neutrins solars poden ser explicades inclús per a angles de mescla
grans (θ > π/4), sempre que els valors dels paràmetres de NSI siguen lo suficientment
grans també. Per tant, malgrat que les NSI tenen un paper secundari en l’explicació
dels fluxos de neutrins solars, encara tenen influència a l’hora de determinar una solució.
Aixı́ doncs, una qüestió sorgeix immediatament: com és possible eliminar la degeneració
de la solució LMA? Aquest és un repte per experiments de neutrins solars com Borexino, SNO o Super-Kamiokande i per a noves propostes com DUNE [221] LENA [293] o
altres LBNE (Long Baseline Neutrino Experiments) [188, 294]. Aixı́ mateix, els experiments de neutrins atmosfèrics i de laboratori com Hyper-Kamiokande [206, 207], IceCube
i DeepCore [295, 296] seran de gran ajuda per tal de descartar la regió “dark”.
Seguint amb l’estudi dels efectes de les NSI sobre els experiments de neutrins, una nova
anàlisi dels resultats de NuTeV s’ha dut a terme en aquesta tesi. Com és ben conegut,
NuTeV presentà una discrepància amb els paràmetres del Model Estàndard, sent utilitzada
com a possible indici de NSI. Però, ja que certes incerteses relacionades amb QCD foren
omeses, hem realitzat una reanàlisi dels resultats de NuTeV en aquesta tesi. D’aquesta
discussió hem obtingut diversos lı́mits en els acoblaments de NSI del neutrı́ muònic amb
quarks (εqµα ). La nostra conclusió ha estat que, encara que aquests paràmetres de NSI
són més menuts que aquells que impliquen neutrins del electró i del tau, no ho són tant
com s’esperava.
Aquests estudis ens han permet obtindre lı́mits en les components vectorial i axial
dels paràmetres de NSI en processos d’interacció de neutrins amb quarks (εqV,A
αα ). En concret, els acoblaments relacionats amb els neutrins de l’electró i el tau han sigut aconseguits
combinant dades d’experiments solars, de reactor (KamLAND) i d’accelerador (CHARM);
mentre que els paràmetres que impliquen neutrins muònics han sigut obtinguts utilitzant
els resultats de la reanàlisi de NuTeV juntament amb dades d’accelerador (CHARM i
CDHS) i d’atmosfèric (Super-Kamiokande). Els resultats més rellevants es troben recopilats a la Taula 6.1 al 90% de C.L.
En l’últim capı́tol d’aquesta tesi hem assumit l’existència de leptons pesats i neutres
(NHL) i hem estudiat les implicacions de la seva existència. Aquestes partı́cules pesades
apareixen en la gran majoria de models estesos com a partı́cules d’intercanvi dins del
mecanisme que genera la massa del neutrı́. Com a conseqüència, les NSI podrien aparèixer
145
directament o de forma indirecta, producte de la no unitarietat de la matriu de mescla
dels neutrins lleugers, com mostrem en les Eqs. (6.1). Aquests models més enllà del SM
normalment impliquen una gran quantitat de paràmetres dificultant el treball amb ells.
Per tal de facilitar l’estudi d’aquests models, hem presentat un formalisme (independent
del model) que ens proporciona una forma molt útil de separar la matriu de mescla dels
neutrins lleugers (N 3×3 ) en dos: l’estàndard (U SM ≡ U 3×3 ) i un altra que concentra tota
la informació de nova fı́sica (N N P ). Aquest formalisme ha sigut aplicat a:
• obtindre lı́mits als paràmetres αij a partir de tests d’universalitat,
• descriure les probabilitats d’oscil·lació, observant efectes interessants com l’oscil·lació
de neutrins a distància zero i l’aparició d’una nova fase IN P la qual podria variar
les probabilitats d’oscil·lació,
• descriure esquemes amb (3 + 1) i (3 + 3) neutrins 2 .
Aquestes aplicacions posen de manifest la simplicitat d’aquest mètode. Com a un projecte
futur, seria molt convenient escriure els acoblaments de NSI en funció dels paràmetres
αij , per exemple:
|α21 ||α31 |
|α21 ||α31 |
, εeR
,
εeL
ee = −gR
ee = −gL
|α32 |α22
|α32 |α22
on s’ha considerat un escenari amb només un neutrı́ extra.
Buscant la completesa d’aquest treball, s’ha realitzat una compilació del valor dels
paràmetres d’acoblament |Sαj |2 dels neutrins pesats amb els lleugers a partir de diferent
experiments de laboratori. Ens agradaria assenyalar que estudiar aquests lı́mits podria
ser de gran utilitat per tal de dilucidar quin és el mecanisme generador de la massa del
neutrı́ aixı́ com el valor de l’escala d’energia de nova fı́sica. Però no és possible trobar
cotes determinants en aquests paràmetres, aixı́ doncs s’han de considerar noves propostes
experimentals, com la ja mencionada de LENA. Com a exemple, una estimació de la
sensitivitat de LENA a la no unitarietat de la matriu de mescla, en funció de α11 , pot
trobar-se a la Fig. 6.1.
Per resumir aquest treball en un parell de frases podrı́em dir: Els Neutrins han obert
la porta a la nova fı́sica, que implica efectes no estàndard i partı́cules extra entre altres
caracterı́stiques. Davant aquestes perspectives, deurı́em d’estar ben atents al que ens
envolta perquè una etapa molt interessant en fı́sica de neutrins ens espera.
2
Veure apèndix B.
146
Chapter 6. Conclusions
Appendix A
NSI in a two-neutrino scheme
It is obvious that working with less parameters simplifies calculations and makes easier to
understand the consequences derived from them. This is obtained when we pass from a
description with three neutrinos to a two-neutrino scheme. The benefits of this approximation were already discussed in Sec. 1.5.2 and it is usually done because reproduces quite
accurately the scenario to study. In fact, it was carried out in Sec. 3.1.2, in particular
when the NSI matrix was defined in Eq. (3.3). Thus, it is worth showing step by step
how this simple calculation is performed, in order to provide rigor to our results obtained
in chapter 3.
In a three-neutrino scheme, the NSI matrix is defined as a 3 × 3 matrix containing all
the NSI couplings:


εee εeµ εeτ


(A.1)
 εµe εµµ εµτ  .
ετ e ετ µ ετ τ
This matrix can be rotated in the µ − τ plane thanks to the ω23 rotation matrix 1 :




1
0
0
εee εeµ εeτ
1
0
0




 0 cos θ23 − sin θ23   εµe εµµ εµτ   0 cos θ23 sin θ23  ,
0 sin θ23 cos θ23
ετ e ετ µ ετ τ
0 − sin θ23 cos θ23
1
(A.2)
Here, we are using the Okubo’s notation [46] as in the Appendix B.
147
148
Appendix A. NSI in a two-neutrino scheme
getting explicitly

εee
c23 εeµ − s23 εeτ
s23 εeµ + c23 εeτ


 c23 εµe − s23 ετ e c223 εµµ − s23 c23 ετ µ − s23 c23 εµτ + s223 ετ τ c23 s23 εµµ − s223 ετ µ + c223 εµτ − s23 c23 ετ τ ,
s23 εµe + c23 ετ e c223 ετ µ + s23 c23 εµµ − s223 εµτ − s23 c23 ετ τ s223 εµµ + s23 c23 ετ µ + s23 c23 εµτ + c223 ετ τ

(A.3)
with c23 = cos θ23 and s23 = sin θ23 . In this point we make the approximation from three to
two neutrinos, neglecting NSI couplings related with muon-neutrinos (εµµ = εµe = εµτ =
0) 2 and truncating the matrix in Eq. (A.3) to a 2 × 2 matrix. After these considerations,
we get the following expression for the NSI hamiltonian:
HN SI
√
= ± 2GF Nf
εee
−s23 εeτ
−s23 εeτ s223 ετ τ
!
.
(A.4)
As the Schrödinger equation does not change if an extra element is introduced in the
diagonal entries of the hamiltonian, we can rewrite Eq. (A.4) as
HN SI
√
= ± 2GF Nf
0
−s23 εeτ
−s23 εeτ
s223 ετ τ − εee
!
,
(A.5)
where it is possible identify
ε = − sin θ23 εeτ ,
ε0 = sin2 θ23 ετ τ − εee ,
(A.6)
as we wanted to prove.
2
We neglect these parameters because we do not take into account muon-neutrinos in the scenario
considered in chapter 3. Obviously, we could cancel other NSI couplings if the situation requires it.
Appendix B
Neutrino mixing and heavy
isosinglets
Neutral heavy leptons are included in several extensions of the Standard Model, giving
rise to new interactions (NSI) or non-unitary effects of the lepton mixing matrix, as we
have already discussed in Secs. 1.4 and 2.2. In these models the neutrino mixing matrix
includes extra terms involving neutral heavy leptons [9], with effects on the light neutrino
sector since the light mixing matrix 1 is not longer unitary. In this appendix we will show
explicitly how the light mixing matrix loses its unitarity and we will develop, step by step,
the α-formalism till we arrive to the αij expressions, Eqs. (5.4), (5.6) and (5.5).
Let us start constructing the mixing matrix for n neutrinos U n×n from Okubo’s notation [46]:
U n×n = ωn−1 n ωn−2 n . . . ω1 n ωn−2 n−1 ωn−3 n−1 . . . ω1 n−1 . . . ω2 3 ω1 3 ω1 2 ,
(B.1)
where each ωij (i < j) represents the common complex rotation matrix in the ij plane [9,
45]:


c13
0 e−iφ13 s13


ω13 = 
(B.2)
0
1
0
,
−eiφ13 s13 0
c13
with sij = sin θij and cij = cos θij . In a more compact form, it can be written as:
(ωij )αβ = δαβ
1
q
1 − δαi δβj s2ij − δαj δβi s2ij + ηij δαi δβj + η̄ij δαj δβi ,
(B.3)
We call light mixing matrix, to the mixing matrix corresponding to the three light neutrinos.
149
150
Appendix B. Neutrino mixing and heavy isosinglets
where i < j and s2ij = sin2 θij , ηij = e−iφij sin θij and η̄ij = −eiφij sin θij . For a general
treatment, Eq. (B.2) can be expressed in a n-neutrino case as:










ωij = 









···
1 0
0 ···
0 1
..
.
cij · · · 0 · · ·
.. . .
.
.
0
..
.
..
.
0
1
ηij
..
.
...
0
..
.
η̄ij · · ·
0 ···
cij
···
0 ···

0
.. 
. 







.




.. 
. 


1 0 
0 1
(B.4)
As we will see later, it is very useful to separate Eq. (B.1) in a product of two matrices,
U n×n = U n−N U N ,
(B.5)
U N = ωN −1 N ωN −2 N . . . ω1 N ,
(B.6)
U n−N = ωn−1 n ωn−2 n . . . ω1 n ωn−1 n−1 ωn−2 n−1 . . . ω1 N +1 .
(B.7)
with
In a matrix form, Eq. (B.5) can be written as:



















α11
α21
..
.
0
α22
···
...
0
..
.
..
0
.
αN 1 · · ·
···
···
αN N
···
V0
···
..
.
..
.
..
.
..
.
..
. ···
..
.
..
.
..
.






S





··· ··· 






T

N
U11
N
U12
N
U1N
···

..
 N
N
 U21 U22
.
 .
..
 .
.
 .

 UN · · ·
UNNN
 N1

 ··· ··· ··· ···






0

..
.
..
.
..
. 0
..
.
..
. ···
..
.
..
. I
..
.










,
··· 







(B.8)
which will be very convenient in order to separate new physics and the Standard Model.
The standard neutrino mixing matrix U 3×3 (Eq. (1.120) or (1.121)) is unitary in the
151
usual description of three neutrinos. But if we consider an extended lepton sector with
n neutrinos, it is a non-unitary (in general) submatrix of the total mixing matrix U n×n
given by Eq. (B.1). Using the property represented by Eq. (B.5), the extended mixing
matrix U n×n can be divided in a more appropriate way as a matrix U N P , which contains
the new physics information, and the usual lepton mixing matrix U SM ≡ U 3×3 :
U n×n = U N P U SM ,
(B.9)
defining each matrix as
U N P = ωn−1 n ωn−2 n . . . ω3 n ω2 n ω1 n ωn−2 n−1 . . . ω3 n−1 ω2 n−1 ω1 n−1 . . . ω3 4 ω2 4 ω1 4 , (B.10)
U SM = ω2 3 ω1 3 ω1 2 .
(B.11)
The total mixing matrix for n neutrinos U n×n can be also expressed in a description
of four blocks as [241]
!
N
S
U n×n =
,
(B.12)
V T
where N ≡ N 3×3 stands for the mixing matrix of the three light neutrinos, which includes
the standard oscillation parameters together with those related with the extra heavy
neutrinos, breaking the unitarity given in a three-neutrino scenario. The terms of this
submatrix N are those that can be observed in oscillation experiments, that is why N is
considered the most important part of U n×n . The submatrix N can be also decomposed,
as in Eqs. (B.5) and (B.9), in a more convenient form as

N = N N P U 3×3

α11 0
0


=  α21 α22 0  U 3×3 .
α31 α32 α33
(B.13)
This division is characterized by the zero triangle submatrix, that depends on the arbitrary
order of the ωij matrices product.
As we have already mentioned, it is interesting to show how the αij components of
the N matrix can be obtained. In order to explain this calculation more accurately, we
should denote that ωij ωkl commutes when i 6= k, l and j 6= k, l; so the new physics part
152
Appendix B. Neutrino mixing and heavy isosinglets
of the total mixing matrix, Eq. (B.10), can be split as follows:
U N P = ωn−1 n ωn−2 n . . . ω4 n ωn−2 n−1 . . . ω4 n−1 . . . ω4 5 ×
ω3 n ω2 n ω1 n ω3 n−1 ω2 n−1 ω1 n−1 . . . ω3 4 ω2 4 ω1 4 .
(B.14)
This decomposition of Eq. (B.10) is very suitable because, as it can be observed, the first
line of Eq. (B.14) has no influence in the relevant submatrix N . However, the second
line components introduce the new physics in the light mixing matrix N , being a set of
products of three matrices in the form ω3j ω2j ω1j . Each product will be identified with a
parameter αj with the following explicit form:
..
.
..
.
..
.
···
..
.
..
.
..
.

j
α = ω3j ω2j ω1j
c1j
0
0

 η η̄
c2j
0
 2j 1j

 η3j c2j η̄1j η3j η̄2j c3j


= 
···
···
···


0
0
0


 c c η̄
 3j 2j 1j c3j η̄2j η̄3j
0








= 






j
α11
0
0
j
j
α21
α22
0
0
j
j
j
α31
α32
α33
··· ··· ···
0
0
0
j
αj1
j
α2j
j
αj3
0
0
0
0
..
.
..
.
..
.
···
..
.
..
.
..
.
0
η1j
0
η2j c1j
0 η3j c2j c1j
···
···
0
j
α1j
0
j
α2j
j
0 α3j
··· ···
I
0
0
j
αjj
0
0
I
0
0
c3j c2j c1j
0
0
0
0


0 


0 


··· 


0 

0 

I


0 


0 


···  .


0 

0 

(B.15)
I
We can see that each αj parameter contains the new physics information introduced by
the i-neutral heavy lepton, but the relevant parameters for N N P are those coming from
the 3 × 3 submatrix located in the upper left side, which presents a triangle submatrix as
we expected. After solving the multiplication given by αn αn−1 · · · α5 α4 , we can find the
153
explicit value of the αij entries in Eq. (B.13):
n
n−1 n−2
4
α11 = α11
α11
α11 · · · α11
= c1 n c 1n−1 c1 n−2 . . . c14 ,
n
n−1 n−2
4
α22 = α22
α22
α22 · · · α22
= c2 n c 2n−1 c2 n−2 . . . c24 ,
(B.16)
n
n−1 n−2
4
α33 = α33
α33
α33 · · · α33
= c3 n c 3n−1 c3 n−2 . . . c34
stand for the diagonal elements αii , while the off-diagonal parameters αij are expressed
as follows:
α21
=
n
n−1
4
n
n−1
4
n
n−1 n−2
4
α21
α11
· · · α11
+ α22
α21
· · · α11
+ · · · + α22
α22
α22 · · · α21
,
α32
=
n
n−1
4
n
n−1
4
n
n−1 n−2
4
α32
α22
· · · α22
+ α33
α32
· · · α22
+ · · · + α33
α33
α33 · · · α32
,
α31
=
n
n−1
4
n
n−1
4
n
n−1 n−2
4
α31
α11
· · · α11
+ α33
α31
· · · α11
+ · · · + α33
α33
α33 · · · α31
+
n
n−1 n−2
4
n−1 n−2
4
n−1 n−2
4
α32
( α21
α11 · · · α11
+ α22
α21 · · · α11
+ · · · + α22
α22 · · · α21
)
+
n
n−1
n−2 n−3
4
n−2 n−3
4
α33
α32
( α21
α11 · · · α11
+ · · · + α22
α22 · · · α21
) + ···
+
n
n−1 n−2
n−3 n−4
4
n−3 n−4
4
α33
α33
α32 ( α21
α11 · · · α11
+ · · · + α22
α22 · · · α21
) + ···
+
n
n−1 n−2
5
4
α33
α33
α33 · · · α32
α21
,
(B.17)
or using a more explicit notation,
α21 = c2 n c 2n−1 . . . c2 5 η24 η̄14 + c2 n . . . c2 6 η25 η̄15 c14 + . . .
+ η2n η̄1n c1n−1 c1n−2 . . . c14 ,
α32 = c3 n c 3n−1 . . . c3 5 η34 η̄24 + c3 n . . . c3 6 η35 η̄25 c24 + . . .
+ η3n η̄2n c2n−1 c2n−2 . . . c24 ,
α31 = c3 n c 3n−1 . . . c3 5 η34 c2 4 η̄14 + c3 n . . . c3 6 η35 c2 5 η̄15 c14 + . . .
+ η3n c2 n η̄1n c1n−1 c1n−2 . . . c14 + c3 n c 3n−1 . . . c3 5 η35 η̄25 η24 η̄14
+
c3 n . . . c3 6 η36 η̄26 c2 5 η24 η̄14 + . . . + η3n η̄2n η2n−1 η̄1n−1 c1n−2 . . . c14 .
(B.18)
To summarize this appendix, we may say that the αij parameters agglutinate all the
new physics information. These parameters along with the known entries of the standard
mixing matrix U 3×3 construct the light mixing matrix N in Eq. (B.13), whose terms could
be observed experimentally. Notice that if there is no extra neutrino, αij = 0 and the
Standard Model description is recovered.
The form of N N P , with an upper right triangular submatrix, has been chosen on
154
Appendix B. Neutrino mixing and heavy isosinglets
purpose. This structure gives rise to the most convenient form of the mixing matrix N
in order to study the most interesting scenarios from a phenomenological point of view.
Therefore, if another order of the ωij products is chosen, we can find the zeros located at
different entries of N N P , being interesting to analyze a particular situation.
Losing the unitarity
In the Standard Model, the neutrino sector is described with three neutrinos (and their
corresponding antineutrinos) as we have discussed in Sec. 1.5.1. The mixing matrix connecting flavor and mass eigenstates is a 3 × 3 unitary matrix (Eq. (1.119)) with the
following generic form:

U SM ≡ U 3×3

Ue1 Ue2 Ue3


=  Uµ1 Uµ2 Uµ3  .
Uτ 1 Uτ 2 Uτ 3
(B.19)
In a most general case, where n neutrinos are taken into account (including, of course, the
three light neutrinos among them) the unitary mixing matrix have n rows and n columns:
N 3×3
S 3×n−3
T n−3×3 V n−3×n−3
U n×n =
!
.
(B.20)
Therefore, a mixing matrix including all neutrinos is unitary, Eq. (1.93). On the other
hand, in models with extended lepton sector, the light mixing matrix N 3×3 follows from
the truncation of U n×n , so it has not to be unitary in general. It has the following generic
structure in our α-formalism:


α11 Ue1
α11 Ue2
α11 Ue3


α21 Ue1 + α22 Uµ1
α21 Ue2 + α22 Uµ2
α21 Ue3 + α22 Uµ3
.

α31 Ue1 + α32 Uµ1 + α33 Uτ 1 α31 Ue2 + α32 Uµ2 + α33 Uτ 2 α31 Ue3 + α32 Uµ3 + α33 Uτ 3
(B.21)
The inclusion of the αij parameters, where the new physics is encoded, breaks the unitarity:
N N † 6= 1 ,
N † N 6= 1 ,
X
i
∗
Nαi Nβi
6= δαβ ,
X
α
∗
Nαi Nαj
6= δij .
(B.22)
155
For instance we have:
∗
∗
∗
2
Ne1
Ne1 + Ne2
Ne2 + Ne3
Ne3 = α11
,
∗
∗
∗
∗
Nµ1
Ne1 + Nµ2
Ne2 + Nµ3
Ne3 = α11 α21
,
(B.23)
(B.24)
instead of 1 or 0 as they should be respectively, if the unitarity condition was accomplished.
3 + 1 seesaw scheme with α parameters
In order to illustrate the convenience of the α-formalism we will apply it to a 3+1 scheme,
where an extra right-handed singlet is added to the SM description:
ΨL =
νL
lL
!
, N4 ,
(B.25)
giving rise to the following relation among flavor and mass eigenstates [5],
ναL =
3
X
Nαk νkL + Sα 4 N4cL .
(B.26)
k=1
In this scenario, the complete mixing matrix U 4×4 is given by

N
e1 Ne2 Ne3 Se4
! 

Nµ1 Nµ2 Nµ3 Sµ4 
S 3×1
,
=

N
V 1×1
 τ 1 Nτ 2 Nτ 3 Sτ 4 
T41 T42 T43 V

U 4×4 =
N 3×3
T 1×3
(B.27)
where N 3×3 is again the submatrix related with the standard neutrinos, Eq. (B.13). It is
important to remember that U 4×4 is unitary because includes extra and standard neutrinos, while N 3×3 is not since it comes from the U 4×4 truncation.
The light mixing matrix has the form given in Eq. (B.21), and the αij parameters are
obtained from Eqs. (B.17) and (B.18), being expressed as follows:
α11 = c14 ,
α22 = c24 ,
α33 = c34 ,
α21 = η24 η̄14 ,
α32 = η34 η̄24 ,
α31 = η34 c24 η̄14 .
(B.28)
156
Appendix B. Neutrino mixing and heavy isosinglets
Application to the 3 + 3 seesaw scheme
In general, the lepton sector is enlarged with more than one extra neutrino in order to
produce the neutrino mass. As we have seen in Sec. 1.4, models including three extra
singlets (as standard seesaw) or more (as inverse or linear seesaw) are common in the
literature. In particular, for a 3 + 3 description the full mixing matrix will have the
following structure,
!
N
S
3×3
3×3
U 6×6 =
,
(B.29)
T3×3 V3×3
with these expressions for the αij parameters:
α11 = c16 c15 c14 ,
α22 = c26 c25 c24 ,
α33 = c36 c35 c34 ,
α21 = η26 η̄16 c15 c14 + c26 η25 η̄15 c14 + c26 c25 η24 η̄14 ,
α32 = c36 c35 η34 η̄24 + c36 c35 η̄25 c24 + η36 η̄26 c25 c24 ,
α31 = c36 c35 c34 η34 c24 η̄14 + c36 η35 c24 η̄15 c14 + η36 c26 η̄16 c15 c14
+ c36 η35 η̄25 η24 η̄14 + η36 η̄26 c25 η24 η̄14 + η36 η̄26 η25 η̄15 c14 .
(B.30)
Appendix C
Relevant neutrino experiments
In this thesis, several neutrino experiments have been mentioned and their data have been
used to perform our calculations. Therefore, we consider that it is worth dedicating this
appendix to explain briefly their features.
C.1
Reactor experiments
Reactor neutrino experiments analyze the number of antineutrino events placing neutrino
detectors around nuclear power plants. This study is performed in order to search for
differences between the expected and the observed antineutrino signal, since this could be
a hint of neutrino oscillations.
In this kind of experiments, antineutrinos are detected through inverse β-decay process
on protons: ν̄e + p → e+ + n. In particular, when an antineutrino travels through the
detector it interacts with a proton giving rise to a positron and a neutron. The produced
positron annihilates with an electron generating photons, whereas the neutron is absorbed
by the liquid scintillator in the detector leading to more photons as a signal of the process.
In the Fig. C.1 we can find a diagram of this detection process.
Double Chooz
Double Chooz [63] is a reactor neutrino experiment with the main goal of measuring the
θ13 mixing angle. It is located at the border between France and Belgium near the Chooz
nuclear plant.
Double Chooz uses two detectors of electron-antineutrinos in order to measure the
number and energy of antineutrinos coming from a pair of nuclear reactors. The near
157
158
Appendix C. Relevant neutrino experiments
Figure C.1: Detection process in a reactor experiment [298].
detector is located 400 m away from reactors, while the far detector is 1.1 km away from
the nuclear cores.
In these detectors, the innermost volume is filled with liquid scintillator doped with
gadolinium. A schematic description of the detectors is given in Fig. C.2.
A difference between the number of events detected in near and far detectors would
imply that some electron-antineutrinos have oscillated in their path towards the detector.
The best-fit value of θ13 measured in Ref. [63] is
sin2 2θ13 = 0.090+0.032
−0.029 .
(C.1)
Daya Bay
The Daya Bay neutrino experiment [64, 65] is a neutrino oscillation experiment. It is
located in Daya Bay (China) where we find two nuclear power plants: Daya Bay NPP
and Ling Ao NPP (Fig. C.3), having 6 nuclear reactors in total.
In the Daya Bay Complex we find three experimental underground halls (EH) where
8 antineutrinos detectors (AD) are placed. Each detector is filled with a liquid scintillator
doped with gadolinium. They are protected against ambient radiation with high-purity
water. A detailed description of these detectors can be found in Refs. [64, 65, 299, 300].
As in Double Chooz, antineutrinos are detected via the inverse β-decay process (Fig. C.1),
showing oscillation events if the ratio between the number of expected and observed antineutrinos is smaller than 1. In a three-neutrino framework, the following results are
C.1. Reactor experiments
Figure C.2: Schematic view of the Double Chooz detector [63].
Figure C.3: Schematic view of Daya Bay experiment. [301].
159
160
Appendix C. Relevant neutrino experiments
derived from Daya Bay data [65]:
sin2 2θ13 = 0.090+0.008
−0.009 ,
−3
|∆m2ee | = (2.59+0.19
eV2 .
−0.20 ) × 10
(C.2)
RENO
RENO (Reactor Experiment for Neutrino Oscillations) [67, 302] is a short baseline reactor
experiment placed at South Korea. It has two identical antineutrino detectors, situated at
294 m and 1383 m from the Hanbit nuclear power plant, composed by six reactors. Each
detector consists of four coaxial layers of cylindrical vessels containing different liquids
depending on the layer purposes (target, gamma-catcher, buffer and veto).
Again, comparing the number of events in the near and far detector, the RENO
Collaboration found the following value for θ13 [302]:
sin2 2θ13 = 0.100 ± 0.010 (stat) ± 0.015 (syst) .
(C.3)
KamLAND
The Kamioka Liquid Scintillator Antineutrino Detector (KamLAND) [37, 303] is an experiment that was located at the Kamioka mine in Japan. The antineutrino flux comes
from 53 nuclear reactors surrounding the zone. These reactors are placed, on average, at
180 km away from the Kamioka mine, making it sensitive to the neutrino mixing related
with the Large Mixing Angle (LMA), the solar neutrino solution with ∆m2 ∼ 10−5 eV2 .
The KamLAND detector consists in a 18 m sphere of stainless steel divided in several
layers. The outer layer is plenty of photomultiplier tubes, while the second inner layer
is a 13 m balloon filled with a scintillating liquid (mineral oil, benzene and fluorescent
chemicals). Finally, a cylindrical water Cherenkov detector is placed around the containment vessel with two purposes: muon veto counter and protect the detector from cosmic
rays and radioactivity.
The KamLAND Collaboration found the following best-fit values for the mixing parameters [37]:
+0.15
−5
eV2 ,
∆m221 = 7.58+0.14
−0.013 (stat)−0.15 (syst) × 10
+0.10
tan2 θ12 = 0.56+0.10
−0.07 (stat)−0.86 (syst) .
(C.4)
C.2. Solar experiments
161
Figure C.4: The detection processes in the SNO experiment [304]. From left to right, the
panels show CC, NC and neutrino-electron scattering respectively.
Combining the results in Eq. (C.4) with solar neutrino data, it was obtained
−5
∆m221 = 7.59+0.21
eV2 ,
−0.21 × 10
C.2
tan2 θ12 = 0.47+0.06
−0.05 .
(C.5)
Solar experiments
Solar neutrino experiments study neutrinos produced in the Sun. They measure the
number of neutrino events in a detector comparing it with the SSM prediction in order to
constrain neutrino oscillation parameters. As an example of detection processes in solar
neutrino experiments, we show those produced in SNO experiment in Fig. C.4.
SNO
The Sudbury Neutrino Observatory (SNO) [35, 36, 70, 71, 168, 169] was placed in the
Creighton mine, in Sudbury (Canada), 2070 m underground.
The experiment was built to detect the energy and direction of 8 B neutrino flux produced in the Sun, distinguishing between electron-neutrinos and other flavors.
Neutrinos coming from the Sun interact with the detector via charged-current weak
interactions, neutral-current weak interactions and neutrino-electron elastic scattering,
as we can see in Fig. C.4. These interactions are registered by 9600 photomultipliers
distributed around the detector (Fig. C.5).
The SNO Collaboration reported an evidence of flavor changing in the solar neutrino
flux [35, 36]. Along with this, the SNO experiment measured a 8 B neutrino flux of [71]
+0.107
Φ8 B = 5.046+0.159
−0.152 (stat)−0.123 (syst) ,
(C.6)
162
Appendix C. Relevant neutrino experiments
Figure C.5: Picture of SNO detector [304].
leading, in combination with all other solar data and KamLAND, to the following best-fit
values for θ12 and ∆m221 :
θ12 (o ) = 34.06+1.16
−0.84 ,
−5
∆m221 = 7.59+0.20
eV2 .
−0.21 × 10
(C.7)
Super-Kamiokande
The Super-Kamiokande experiment [68, 69, 75, 167] is a collaboration between 30 institutions of several countries (Japan, United States, Korea, China Poland and Spain).
The detector is placed 1000 m underground in the Kamioka mine in Japan. This experiment observes neutrinos stemming from the Sun, atmospheric neutrinos as well as
neutrinos produced artificially. Neutrino oscillations were observed for the first time in
Super-Kamiokande in 1998.
Super-Kamiokande is a large water Cherenkov detector (Fig. C.6), which consists in
a spherical stainless steel tank of 39 m of diameter, filled with 50.000 tons of ultra-pure
water. 13.000 photomultipliers are installed in the inner surface of the tank wall, in order
to register the Cherenkov radiation produced after neutrino scattering with electrons
C.2. Solar experiments
163
Figure C.6: Schematic image of the Super-Kamiokande detector [305].
(Fig. C.4).
From a global analysis including all current solar neutrino data (incorporating even
SK-IV) together with KamLAND results, the Super-Kamiokande Collaboration published
the following values for the oscillation parameters [69]:
−5
∆m221 = 7.49+0.19
eV2 ,
−0.17 × 10
sin2 θ12 = 0.305 ± 0.013 .
(C.8)
Borexino
Borexino [74, 170, 171] is an experiment which studies low-energy neutrinos coming from
the Sun, with the particular goal of measuring 7 Be solar neutrino flux and compare it
with the Standard Model prediction. It is located at the Laboratori Nazionali del Gran
Sasso near of L’Aquila (Italy).
The detector is a liquid scintillator, where the neutrino interactions are registered
by 2200 photomultipliers that surround it. The liquid scintillator is inside a stainless
steel sphere, being protected by a water tank. For a schematic view of the detector, see
Fig. C.7.
164
Appendix C. Relevant neutrino experiments
Figure C.7: Schematic view of the Borexino detector [306].
From the analysis of Borexino data, it was found that [74]
−5
∆m221 = 5.2+1.5
eV2 ,
−0.24 × 10
tan2 θ12 = 0.468+0.039
−0.030 .
(C.9)
After including the KamLAND results, the oscillation parameters became
−5
∆m221 = 7.50+0.16
eV2 ,
−0.9 × 10
C.3
tan2 θ12 = 0.457+0.033
−0.025 .
(C.10)
Accelerator experiments
Accelerator neutrino experiments are experimental devices using neutrinos produced at
accelerators as source. In particular, neutrinos come from hadron decays as, for instance,
K + → e+ ν ,
K + → µ+ ν ,
π + → e+ ν ,
π + → µ+ ν .
(C.11)
C.3. Accelerator experiments
165
Figure C.8: Schematic view of the CHARM detector [160].
CHARM
CHARM (Cern Hamburg Rome Moscow) [160, 194, 307] was a collaboration which tried
to shed some light on some aspects of electroweak theory as the Weinberg angle or the
universality of the weak couplings. The experiment measured the ratios
Re =
σ(νe N → νe X)
,
νe N → e X
Rµ =
σ(νµ N → νµ X)
,
νµ N → µ X
(C.12)
using accelerator neutrinos produced in SPS (CERN). In particular, electron-neutrinos
were obtained from K 0 decay, whereas muon-neutrinos were produced at π decay.
The CHARM detector was a calorimeter surrounded by a muon spectrometer. This
calorimeter consisted in 72 marble plates connected through scintillating plates. An image
of this detector is given in Fig. C.8. The calorimeter recorded neutrino events produced
from neutral-current or charged-current weak interactions.
NuTeV
NuTeV [162] was a detector designed to observe the interactions of neutrinos produced
in the Tevatron accelerator at Fermilab, after the collision of 800 GeV protons on a BeO
target. The detector was built with three types of layers (Fig. C.9).
The results obtained by NuTeV were quite controversial because they were far from
the SM predictions, as we discussed in chapter 4.
166
Appendix C. Relevant neutrino experiments
Figure C.9: The left panel shows neutrino production at Fermilab. The right: scheme of
the NuTeV detector [308].
Figure C.10: Signals registered in the NuTeV detector: CC signal and the NC signal
respectively [308].
C.4
Experiments giving constraints on NHL
π → e ν (TRIUMF)
This experiment [265, 266] tested the universality from studying the branching ratio of
π → e ν with respect to the common decay π → µ ν obtaining [265]:
Rπeν =
Γ(π → e ν + π → e ν γ)
= 1.2265 ± 0.0034(stat) ± 0.0044(syst) × 10−4 . (C.13)
Γ(π → µ ν + π → µ ν γ)
The experiment was performed in the M13 channel at TRIUMF, using a π + beam
of momentum P = 83 MeV/c and ∆P/P = 1. The incoming beam was detected in a
scintillator and stopped close to the target counter with a rate of 7 × 104 s−1 . Then, the
produced positrons were detected with two 203 mm×203 mm wire chambers, before being
analyzed in a 460 mm diam × 510 mm long NaI crystals, ”TINA.”
The mixing with extra heavy neutrinos should produce additional peaks in the positron
energy spectrum from decays like π + → e+ ν and K + → e+ ν, leading to the following
C.4. Experiments giving constraints on NHL
167
ratio which can be related to a heavy state [266]:
Rei =
Γ(π → e νi )
= |Uei |2 ρe ,
Γ(π → e ν1 )
(C.14)
with ν1 being a massless neutrino.
PS191
PS191 [267] was an experiment located at CERN, which was expressly designed to search
for neutrino decays in a low-energy neutrino beam. The apparatus consisted in a ∼ 10
m long decay volume, built with eight pairs of flash chambers and a shower detector,
including also a scintillator hodoscope for triggering purposes.
Several channels were studied, as K + → e+ ν and νH → e+ e− ν, being used to find
bounds on the matrix elements |Ue |2 and |Ue∗ Uµ | as a function of mH .
NA3
NA3 [268] was a spectrometer used in SPS at CERN in order to search new and long-lived
particles. The experiment used a beam of 300 GeV/c of π − , colliding on an hadron iron
absorber of 2 m long.
The device had to be able to detect neutrinos heavy enough to decay in a final states,
which could be: e+ e− ν, µ+ µ− ν, π + e− , µ+ e− ν or µ+ µ− . Such heavy neutrinos may be
produced in rare decays of π, K, D, F or B mesons. The branching ratio of these mesons
decaying into a heavy a neutrino νh has the following expression related with the mixing
matrix element Uih :
B.R. = c |Uih | ,
(C.15)
with c = 0.1 × 10−3 depending on de mass of νh .
CHARM
In the CHARM experiment [269, 286] (described above), the heavy neutrino production
was searched in decays of the charmed D meson. Heavy neutrinos should have e+ e− νe ,
µ+ e− νe , e+ µ− νµ or µ+ µ− νµ as decaying signature, but no evidence was found.
This experiment was prepared to detect decays of neutrinos with masses in the range
168
Appendix C. Relevant neutrino experiments
0.5 − 2.8 GeV, finding two upper limits [269]:
|Uei |2 |Uµi |2 < 10−7 for masses around 1.5 GeV ,
(C.16)
|Uµi |2 < 3 × 10−4 for masses around 2.5 GeV .
(C.17)
DELPHI
DELPHI [231] was a detector of LEP, collecting data from 3.3 × 106 hadronic Z 0 decays
(e+ e− → Z 0 → νH ν̄) coming from a total sample of 12.3 × 106 events. From the data
analysis, an upper limit was obtained [231],
0
0
2
B.R.(Z → νH ν̄) = B.R.(Z → ν ν̄) |U |
m2νH
1− 2
mZ 0
1 m2νH
1+
2 m2Z 0
< 1.3×10−6 (C.18)
at 95% C.L., leading to
|U |2 < 2.1 × 10−5 .
(C.19)
L3
L3 [271] was a device of LEP, which was composed by a central vertex chamber (TEC)
with inner radius of 9 cm and outer of 47 cm, a high resolution electromagnetic calorimeter
of BGO crystals, a ring of scintillation counters, an uranium and brass hadron calorimeter
and a precise muon chamber system. These detectors were placed around a 12 m diameter
magnet which produced an uniform field of 0.5 Tesla along the beam axis.
As in DELPHI, the search was performed through Z 0 decay (e+ e− → Z 0 → νH ν̄),
obtaining the following values at 95% C.L.:
B.R.(e+ e− → Z 0 → νH ν̄) < 3 × 10−5 ,
|U |2 < 10−4 .
(C.20)
(C.21)
K → µ ν (KEK)
As we have already said, a discrete muon peak in the muon momentum spectrum produced
in K + decays could imply a heavy neutrino emission. This special production could be
possible due to the mixing of heavy neutrinos with the light ones.
In order to look for the emission of a heavy neutrino, a high resolution magnetic
spectrograph was built at the K3 beam channel of the 12 GeV proton synchrotron at the
C.4. Experiments giving constraints on NHL
169
National Laboratory for High Energy Physics (KEK) [278, 279]. It used a 550 MeV/c
K + beam crossing through ten layers of plastic scintillators (seven 8 × 20 × 0.6 cm3 , two
8 × 20 × 0.2 cm3 and one 8 × 20 × 0.1 cm3 ). Although, the number of stopped K + was
high (about 5000 per beam) no heavy neutrino was found.
BEBC
BEBC [282] was a bubble chamber filled with a mixture of Ne/H2 , giving rise to a fiducial
mass of 11.5 t. This material was exposed to a neutrino flux produced by 1.9 × 1018
protons colliding on a solid copper target and 0.18 × 1018 protons hitting a laminated
copper target. BEBC was equipped with an External Muon Identifier. The most relevant
results of BEBC are compiled in Ref. [282].
FMMF
FMMF [281] was a detector designed for studying high-energy neutrino interactions. It
was used at Fermilab Tevatron, being placed 1599 m downstream of the neutrino target. It
consisted of two parts; a calorimeter with alternating planes of target material (sand and
steel) interleaved with flash chamber planes and proportional tube planes, and a muon
spectrometer which measured the momentum and polarity of any high-energy muons
coming from the calorimeter.
CHARM-II
CHARM-II [280, 309] was a neutrino detector at CERN. It consisted of a large target
calorimeter equipped with streamer tubes and a muon spectrometer. It was constructed
in order to look for heavy neutrinos in the muon-neutrino-nucleon scattering and the
subsequent decay into µ+ µ− νµ .
The analysis was based on 2 × 107 NC neutrino events collected between 1987-1991,
searching heavy neutrinos in the mass range 0.3 − 2.4 GeV/c2 . The best limit for the
mixing parameter with muon-neutrino was [280]
|Uµi |2 < 3 × 10−5
(C.22)
for a mass around 2 GeV/c2 .
This new analysis improved by an order of magnitude previous results stemming from
CHARM.
170
Appendix C. Relevant neutrino experiments
NOMAD
A search for heavy neutrinos (νh ) was performed studying the decay D → τ νh at the SPS
proton target. After this decay, another process, with signature νh → ντ e+ e− , could be
produced and detected in NOMAD [285], proving the existence of heavy neutrinos. It
must be said that no evidence was found.
Bibliography
[1] Y. Fukuda et al. Evidence for oscillation of atmospheric neutrinos. Phys. Rev. Lett.,
81:1562–1567, 1998.
[2] S. Pakvasa and J. W. F. Valle. Neutrino properties before and after KamLAND.
Proc. Indian Natl. Sci. Acad., 70A:189–222, 2004.
[3] Jose W. F. Valle and Jorge C. Romao. Neutrinos in high energy and astroparticle physics. Physics textbook, ISBN: 978-3-527-41197-9, 978-3-527-67102-1. WileyVCH, Weinheim, 2015.
[4] J. W. F. Valle. Neutrino physics overview. J. Phys. Conf. Ser., 53:473–505, 2006.
These review lectures were given at Corfu, 2005 and contain extensive references to
the early papers on the seesaw mechanism.
[5] R. N. Mohapatra and P. B. Pal. Massive neutrinos in physics and astrophysics.
Second edition. Physics textbook. World Sci. Lect. Notes Phys. 397 p., 1998.
[6] K. A. Olive et al. Review of Particle Physics. Chin. Phys., C38:090001, 2014.
[7] J. Beringer et al. Review of Particle Physics (RPP). Phys.Rev., D86:010001, 2012.
[8] S. Schael et al. Precision electroweak measurements on the Z resonance. Phys.
Rept., 427:257–454, 2006.
[9] J. Schechter and J.W.F. Valle.
Phys.Rev., D22:2227, 1980.
Neutrino Masses in SU(2) x U(1) Theories.
[10] K. Zuber. Neutrino physics. Physics textbook, ISBN: 9780750307505. Boca Raton:
USA: CRC Pr. 448 p., 2004.
[11] C.W. Kim and A. Pevsner. Neutrinos in Physics and Astrophysics. Physics textbook. Harwood Academic: 429 p., 1993.
171
172
Bibliography
[12] Carlo Giunti and Chung W. Kim. Fundamentals of Neutrino Physics and Astrophysics. Physics textbook. Oxford UK: Univ.Pr. 710 p., 2007.
[13] Walter Grimus. Neutrino Physics - Models for Neutrino Masses and Lepton Mixing.
Proceedings, School on Particle Physics, Gravity and Cosmology. ArXiv, 0612311
[hep-ph]:001, 2006.
[14] Robert Foot, H. Lew, X. G. He, and Girish C. Joshi. Seesaw neutrino masses
induced by a triplet of leptons. Z. Phys., C44:441, 1989.
[15] Peter Minkowski. µ → eγ at a Rate of One Out of 109 Muon Decays? Phys. Lett.,
B67:421, 1977.
[16] Steven Weinberg. Varieties of baryon and lepton nonconservation. Phys. Rev.,
D22:1694, 1980.
[17] J. Schechter and J. W. F. Valle. Neutrino Decay and Spontaneous Violation of
Lepton Number. Phys. Rev., D25:774, 1982.
[18] R. N. Mohapatra and J. W. F. Valle. Neutrino mass and baryon-number nonconservation in superstring models. Phys. Rev., D34:1642, 1986.
[19] M.C. Gonzalez-Garcia and J.W.F. Valle. Fast Decaying Neutrinos and Observable
Flavor Violation in a New Class of Majoron Models. Phys.Lett., B216:360, 1989.
[20] Edward Witten. Symmetry breaking patterns in superstring models. Nucl. Phys.,
B258:75, 1985.
[21] D.V. Forero, S. Morisi, M. Tórtola, and J. W. F. Valle. Lepton flavor violation and
non-unitary lepton mixing in low-scale type-I seesaw. JHEP, 1109:142, 2011.
[22] Michal Malinsky, J.C. Romao, and J.W.F. Valle. Novel supersymmetric SO(10)
seesaw mechanism. Phys.Rev.Lett., 95:161801, 2005.
[23] A. Zee. A theory of lepton number violation, neutrino majorana mass, and oscillation. Phys. Lett., B93:389, 1980.
[24] K. S. Babu. Model of ’calculable’ majorana neutrino masses. Phys. Lett., B203:132,
1988.
[25] J. T. Peltoniemi and J. W. F. Valle. Massive neutrinos and electroweak baryogenesis.
Phys. Lett., B304:147–151, 1993.
Bibliography
173
[26] M. Hirsch and J. W. F. Valle. Supersymmetric origin of neutrino mass. New J.
Phys., 6:76, 2004.
[27] A. Masiero and J. W. F. Valle. A model for spontaneous R parity breaking. Phys.
Lett., B251:273–278, 1990.
[28] J. C. Romao, C. A. Santos, and J. W. F. Valle. How to spontaneously break Rparity. Phys. Lett., B288:311–320, 1992.
[29] J. C. Romao, A. Ioannisian, and J. W. F. Valle. Supersymmetric unification with
radiative breaking of R-parity. Phys. Rev., D55:427–430, 1997.
[30] M. Hirsch et al. Neutrino masses and mixings from supersymmetry with bilinear
R-parity violation: A theory for solar and atmospheric neutrino oscillations. Phys.
Rev., D62:113008, 2000. Err-ibid. D65:119901,2002.
[31] M. A. Diaz et al. Solar neutrino masses and mixing from bilinear R-parity broken
supersymmetry: Analytical versus numerical results. Phys. Rev., D68:013009, 2003.
[32] B. Pontecorvo. Mesonium and anti-mesonium. Sov. Phys. JETP, 6:429, 1957. [Zh.
Eksp. Teor. Fiz.33,549(1957)].
[33] B. Pontecorvo. Inverse beta processes and nonconservation of lepton charge. Sov.
Phys. JETP, 7:172–173, 1958. [Zh. Eksp. Teor. Fiz.34,247(1957)].
[34] T. Futagami et al. Observation of the east - west anisotropy of the atmospheric
neutrino flux. Phys. Rev. Lett., 82:5194–5197, 1999.
[35] Q. R. Ahmad et al. Direct evidence for neutrino flavor transformation from
neutral-current interactions in the Sudbury Neutrino Observatory. Phys. Rev. Lett.,
89:011301, 2002.
[36] Q. R. Ahmad et al. Measurement of day and night neutrino energy spectra at SNO
and constraints on neutrino mixing parameters. Phys. Rev. Lett., 89:011302, 2002.
[37] S. Abe et al. Precision Measurement of Neutrino Oscillation Parameters with KamLAND. Phys.Rev.Lett., 100:221803, 2008.
[38] M. H. Ahn et al. Measurement of Neutrino Oscillation by the K2K Experiment.
Phys. Rev., D74:072003, 2006.
174
Bibliography
[39] K. Abe et al. Indication of Electron Neutrino Appearance from an Acceleratorproduced Off-axis Muon Neutrino Beam. Phys.Rev.Lett., 107:041801, 2011.
[40] K. Abe et al. Observation of Electron Neutrino Appearance in a Muon Neutrino
Beam. Phys. Rev. Lett., 112:061802, 2014.
[41] K. Abe et al. Precise Measurement of the Neutrino Mixing Parameter θ23 from Muon
Neutrino Disappearance in an Off-Axis Beam. Phys. Rev. Lett., 112(18):181801,
2014.
[42] P. Adamson et al. Improved search for muon-neutrino to electron-neutrino oscillations in MINOS. Phys.Rev.Lett., 107:181802, 2011.
[43] P. Adamson et al. Electron neutrino and antineutrino appearance in the full MINOS
data sample. Phys. Rev. Lett., 110(17):171801, 2013.
[44] P. Adamson et al. Measurement of Neutrino and Antineutrino Oscillations Using
Beam and Atmospheric Data in MINOS. Phys.Rev.Lett., 110:251801, 2013.
[45] W. Rodejohann and J. W. F. Valle. Symmetrical Parametrizations of the Lepton
Mixing Matrix. Phys.Rev., D84:073011, 2011.
[46] S. Okubo. Note on Unitary Symmetry in Strong Interaction. II Excited States of
Baryons. Prog.Theor.Phys., 28:24–32, 1962.
[47] Evgeny K. Akhmedov. Neutrino physics. Particle physics. Proceedings, Summer
School, Trieste, Italy, June 21-July 9, 1999. ArXiv, 0001264 [hep-ph], 1999.
[48] L. Wolfenstein. Neutrino Oscillations in Matter. Phys.Rev., D17:2369–2374, 1978.
[49] S. P. Mikheev and A. Yu. Smirnov. Resonant amplification of neutrino oscillations
in matter and solar neutrino spectroscopy. Nuovo Cim., C9:17–26, 1986.
[50] Bruce T. Cleveland et al. Measurement of the solar electron neutrino flux with the
homestake chlorine detector. Astrophys. J., 496:505–526, 1998.
[51] Y. Fukuda et al. Solar neutrino data covering solar cycle 22. Phys. Rev. Lett.,
77:1683–1686, 1996.
[52] W. Hampel et al. Gallex solar neutrino observations: Results for gallex iv. Phys.
Lett., B447:127–133, 1999.
Bibliography
175
[53] J. N. Abdurashitov et al. Measurement of the solar neutrino capture rate by sage
and implications for neutrino oscillations in vacuum. Phys. Rev. Lett., 83:4686,
1999.
[54] Y. Fukuda et al. Measurements of the solar neutrino flux from Super- Kamiokande’s
first 300 days. Phys. Rev. Lett., 81:1158–1162, 1998.
[55] John N. Bahcall. Standard solar models. Nucl. Phys. Proc. Suppl., 77:64–72, 1999.
[56] David Griffiths. Introduction to elementary particles. Physics textbook, ISBN:
9783527406012. Weinheim, Germany: Wiley-VCH (2008) 454 p., 2008.
[57] Jose F. Nieves and Palash B. Pal. Generalized Fierz identities. Am. J. Phys.,
72:1100–1108, 2004.
[58] Samoil M. Bilenky. Introduction to Feynman diagrams and electroweak interactions
physics. Physics textbook. Gif-sur-Yvette, France: Ed. Frontieres (1994) 368 p.,
1995.
[59] D. V. Forero, M. Tórtola, and J. W. F. Valle. Neutrino oscillations refitted. Phys.
Rev., D90(9):093006, 2014.
[60] F. Capozzi, G. L. Fogli, E. Lisi, Marrone A., D. Montanino, and A. Palazzo. Status
of three-neutrino oscillation parameters, circa 2013. Phys. Rev., D89:093018, 2014.
[61] M. C. Gonzalez-Garcia, Michele Maltoni, and Thomas Schwetz. Updated fit to
three neutrino mixing: status of leptonic CP violation. JHEP, 11:052, 2014.
[62] Ed Kearns, editor. Proceedings, 26th International Conference on Neutrino Physics
and Astrophysics (Neutrino 2014), volume 1666, ISBN: 978-0-7354-1313-9, 2015.
[63] Y. Abe et al. Improved measurements of the neutrino mixing angle θ13 with the
Double Chooz detector. JHEP, 10:086, 2014. [Erratum: JHEP02,074(2015)].
[64] F.P. An et al. Observation of electron-antineutrino disappearance at Daya Bay.
Phys.Rev.Lett., 108:171803, 2012.
[65] F. P. An et al. Spectral measurement of electron antineutrino oscillation amplitude
and frequency at Daya Bay. Phys. Rev. Lett., 112:061801, 2014.
[66] Proceedings, 13th International Conference on Topics in Astroparticle and Underground Physics (TAUP 2013). Phys. Procedia, 61:pp.1–838, 2015.
176
Bibliography
[67] J.K. Ahn et al. Observation of Reactor Electron Antineutrino Disappearance in the
RENO Experiment. Phys.Rev.Lett., 108:191802, 2012.
[68] K. Abe et al. Solar neutrino results in Super-Kamiokande-III.
D83:052010, 2011.
Phys.Rev.,
[69] Andrew Renshaw. Solar Neutrino Results from Super-Kamiokande. Phys. Procedia,
61:345–354, 2015.
[70] B. Aharmim et al. An Independent Measurement of the Total Active 8B Solar Neutrino Flux Using an Array of 3He Proportional Counters at the Sudbury Neutrino
Observatory. Phys. Rev. Lett., 101:111301, 2008.
[71] B. Aharmim et al. Low Energy Threshold Analysis of the Phase I and Phase II
Data Sets of the Sudbury Neutrino Observatory. Phys. Rev., C81:055504, 2010.
[72] F. Kaether, W. Hampel, G. Heusser, J. Kiko, and T. Kirsten. Reanalysis of the
GALLEX solar neutrino flux and source experiments. Phys.Lett., B685:47–54, 2010.
[73] J. N. Abdurashitov et al. Measurement of the solar neutrino capture rate with
gallium metal. III: Results for the 2002–2007 data-taking period. Phys. Rev.,
C80:015807, 2009.
[74] G. Bellini, J. Benziger, D. Bick, S. Bonetti, G. Bonfini, et al. Precision measurement
of the 7 Be solar neutrino interaction rate in Borexino. Phys.Rev.Lett., 107:141302,
2011.
[75] J. Hosaka et al. Solar neutrino measurements in Super-Kamiokande-I. Phys. Rev.,
D73:112001, 2006.
[76] J.P. Cravens et al.
Solar neutrino measurements in Super-Kamiokande-II.
Phys.Rev., D78:032002, 2008.
[77] Y. Fukuda et al. Measurement of the flux and zenith-angle distribution of upward
through-going muons by Super-Kamiokande. Phys. Rev. Lett., 82:2644–2648, 1999.
[78] Y. Fukuda et al. Atmospheric muon-neutrino / electron-neutrino ratio in the multigev energy range. Phys. Lett., B335:237–245, 1994.
[79] R. Becker-Szendy et al. Neutrino measurements with the imb detector. Nucl. Phys.
Proc. Suppl., 38:331–336, 1995.
Bibliography
177
[80] W. W. M. Allison et al. The atmospheric neutrino flavor ratio from a 3.9 fiducial
kiloton-year exposure of soudan 2. Phys. Lett., B449:137, 1999.
[81] M. Ambrosio et al. Measurement of the atmospheric neutrino-induced upgoing muon
flux using macro. Phys. Lett., B434:451–457, 1998.
[82] Takaaki Kajita and Yoji Totsuka. Observation of atmospheric neutrinos. Rev. Mod.
Phys., 73:85–118, 2001.
[83] J. W. F. Valle. Resonant oscillations of massless neutrinos in matter. Phys. Lett.,
B199:432, 1987.
[84] O.G. Miranda, M.A. Tórtola, and J.W.F. Valle. Are solar neutrino oscillations
robust? JHEP, 0610:008, 2006.
[85] F. J. Escrihuela, O. G. Miranda, M. A. Tórtola, and J. W. F. Valle. Constraining
nonstandard neutrino-quark interactions with solar, reactor and accelerator data.
Phys. Rev., D80:105009, 2009.
[86] N. Fornengo et al. Probing neutrino non-standard interactions with atmospheric
neutrino data. Phys. Rev., D65:013010, 2002.
[87] S. Davidson, C. Peña-Garay, N. Rius, and A. Santamaria. Present and future bounds
on non-standard neutrino interactions. JHEP, 03:011, 2003.
[88] Joachim Kopp, Manfred Lindner, Toshihiko Ota, and Joe Sato. Non-standard neutrino interactions in reactor and superbeam experiments. Phys. Rev., D77:013007,
2008.
[89] Tommy Ohlsson. Status of non-standard neutrino interactions. Rept. Prog. Phys.,
76:044201, 2013.
[90] O. G. Miranda and H. Nunokawa. Non standard neutrino interactions: current
status and future prospects. New J. Phys., 17(9):095002, 2015.
[91] J. Barranco, O. G. Miranda, and T. I. Rashba. Low energy neutrino experiments
sensitivity to physics beyond the Standard Model. Phys. Rev., D76:073008, 2007.
[92] Murray Gell-Mann, Pierre Ramond, and Richard Slansky. Complex Spinors and
Unified Theories. Conf. Proc., C790927:315–321, 1979.
178
Bibliography
[93] Rabindra N. Mohapatra and Goran Senjanovic. Neutrino Mass and Spontaneous
Parity Violation. Phys. Rev. Lett., 44:912, 1980.
[94] F. J. Escrihuela, D. V. Forero, O. G. Miranda, M. Tórtola, and J. W. F. Valle. On
the description of non-unitary neutrino mixing. Phys. Rev., D92(5):053009, 2015.
[95] Michal Malinsky, Tommy Ohlsson, and He Zhang. Non-Standard Neutrino Interactions from a Triplet Seesaw Model. Phys. Rev., D79:011301, 2009.
[96] M. Drees, R. Godbole, and P. Roy. Theory and phenomenology of sparticles: An
account of four-dimensional N=1 supersymmetry in high energy physics. Physics
book. Hackensack, USA: World Scientific (2004) 555 p, 2004.
[97] Lawrence J. Hall and Mahiko Suzuki. Explicit R parity breaking in supersymmetric
models. Nucl. Phys., B231:419, 1984.
[98] A. Santamaria and J. W. F. Valle. Spontaneous R parity violation in supersymmetry: a model for solar neutrino oscillations. Phys. Lett., B195:423, 1987.
[99] G. G. Ross and J. W. F. Valle. Supersymmetric models without R parity. Phys.
Lett., B151:375, 1985.
[100] Vernon D. Barger, G. F. Giudice, and Tao Han. Some new aspects of supersymmetry
R parity violating interactions. Phys. Rev., D40:2987, 1989.
[101] Marco A. Diaz, Jorge C. Romao, and J. W. F. Valle. Minimal supergravity with
R-parity breaking. Nucl. Phys., B524:23–40, 1998.
[102] Asmaa Abada, Sacha Davidson, and Marta Losada. Neutrino masses and mixings
in the MSSM with soft bilinear R(p) violation. Phys. Rev., D65:075010, 2002.
[103] J. Barranco et al. Constraining Nonstandard Neutrino-Electron Interactions. Phys.
Rev., D77:093014, 2008.
[104] M. M. Guzzo, P. C. de Holanda, and O. L. G. Peres. Effects of non-standard
neutrino interactions on MSW-LMA solution. Phys. Lett., B591:1–6, 2004.
[105] Sven Bergmann and Alex Kagan. Z-induced fcncs and their effects on neutrino
oscillations. Nucl. Phys., B538:368–386, 1999.
[106] M. C. Gonzalez-Garcia et al. Atmospheric neutrino observations and flavor changing
interactions. Phys. Rev. Lett., 82:3202–3205, 1999.
Bibliography
179
[107] Alexander Friedland, Cecilia Lunardini, and Michele Maltoni. Atmospheric neutrinos as probes of neutrino matter interactions. Phys. Rev., D70:111301, 2004.
[108] D. Aristizabal Sierra, M. Hirsch, and S. G. Kovalenko. Leptoquarks: Neutrino
masses and accelerator phenomenology. Phys. Rev., D77:055011, 2008.
[109] Jogesh C. Pati and Abdus Salam. Lepton number as the fourth color. Phys. Rev.,
D10:275–289, 1974.
[110] H. Georgi and S.L. Glashow.
Phys.Rev.Lett., 32:438–441, 1974.
Unity of All Elementary Particle Forces.
[111] Ilja Dorsner and Pavel Fileviez Perez. Unification without supersymmetry: Neutrino
mass, proton decay and light leptoquarks. Nucl. Phys., B723:53–76, 2005.
[112] Ilja Dorsner, Pavel Fileviez Perez, and Ricardo Gonzalez Felipe. Phenomenological
and cosmological aspects of a minimal GUT scenario. Nucl. Phys., B747:312–327,
2006.
[113] Harald Fritzsch and Peter Minkowski. Unified Interactions of Leptons and Hadrons.
Annals Phys., 93:193–266, 1975.
[114] Edward Farhi and Leonard Susskind. Technicolor. Phys. Rept., 74:277, 1981.
[115] Sacha Davidson, David C. Bailey, and Bruce A. Campbell. Model independent
constraints on leptoquarks from rare processes. Z. Phys., C61:613–644, 1994.
[116] Yuval Grossman. Nonstandard neutrino interactions and neutrino oscillation experiments. Phys. Lett., B359:141–147, 1995.
[117] M. C. Gonzalez-Garcia, Y. Grossman, A. Gusso, and Y. Nir. New CP violation in
neutrino oscillations. Phys. Rev., D64:096006, 2001.
[118] Samoil M. Bilenky and C. Giunti. Seesaw type mixing and νµ → ντ oscillations.
Phys. Lett., B300:137–140, 1993.
[119] Davide Meloni, Tommy Ohlsson, Walter Winter, and He Zhang. Non-standard
interactions versus non-unitary lepton flavor mixing at a neutrino factory. JHEP,
04:041, 2010.
[120] Paul Langacker and David London. Lepton Number Violation and Massless
Nonorthogonal Neutrinos. Phys. Rev., D38:907, 1988.
180
Bibliography
[121] M. M. Guzzo, A. Masiero, and S. T. Petcov. On the MSW effect with massless
neutrinos and no mixing in the vacuum. Phys. Lett., B260:152, 1991.
[122] Esteban Roulet. Mikheyev-Smirnov-Wolfenstein effect with flavor-changing neutrino
interactions. Phys. Rev., D44:935–938, 1991.
[123] D. Meloni, T. Ohlsson, and H. Zhang. Exact and Approximate Formulas for Neutrino Mixing and Oscillations with Non-Standard Interactions. JHEP, 04:033, 2009.
[124] F. J. Escrihuela, M. Tórtola, J. W. F. Valle, and O. G. Miranda. Global constraints
on muon-neutrino non-standard interactions. Phys. Rev., D83:093002, 2011.
[125] Sanjib Kumar Agarwalla, Partha Bagchi, David V. Forero, and Mariam Tórtola.
Probing Non-Standard Interactions at Daya Bay. JHEP, 07:060, 2015.
[126] Mattias Blennow and Tommy Ohlsson. Approximative two-flavor framework for
neutrino oscillations with non-standard interactions. Phys. Rev., D78:093002, 2008.
[127] M. C. Gonzalez-Garcia and Michele Maltoni. Atmospheric neutrino oscillations and
new physics. Phys. Rev., D70:033010, 2004.
[128] G. Mitsuka et al. Study of Non-Standard Neutrino Interactions with Atmospheric
Neutrino Data in Super-Kamiokande I and II. Phys. Rev., D84:113008, 2011.
[129] A. M. Gago, M. M. Guzzo, H. Nunokawa, W. J. C. Teves, and R. Zukanovich Funchal. Probing flavor changing neutrino interactions using neutrino beams from a
muon storage ring. Phys. Rev., D64:073003, 2001.
[130] Arman Esmaili and Alexei Yu. Smirnov. Probing Non-Standard Interaction of Neutrinos with IceCube and DeepCore. JHEP, 06:026, 2013.
[131] Alexander Friedland and Cecilia Lunardini. A test of tau neutrino interactions with
atmospheric neutrinos and K2K. Phys. Rev., D72:053009, 2005.
[132] Noriaki Kitazawa, Hiroaki Sugiyama, and Osamu Yasuda. Will MINOS see new
physics? ArXiv: 0606013, [hep-ph], 2006.
[133] Joachim Kopp, Pedro A. N. Machado, and Stephen J. Parke. Interpretation of MINOS data in terms of non-standard neutrino interactions. Phys. Rev., D82:113002,
2010.
Bibliography
181
[134] W. Anthony Mann, Daniel Cherdack, Wojciech Musial, and Tomas Kafka. Apparent
multiple ∆m232 in muon anti-neutrino and muon neutrino survival oscillations from
non-standard interaction matter effect. Phys. Rev., D82:113010, 2010.
[135] Alexander Friedland and Cecilia Lunardini. Two modes of searching for new neutrino interactions at MINOS. Phys. Rev., D74:033012, 2006.
[136] Zeynep Isvan. Search for Non-standard Interactions with the MINOS Experiment.
Particles and fields. Proceedings, Meeting of the Division of the American Physical
Society, DPF 2011, Providence, USA, August 9-13, 2011 . ArXiv, 1110.1900 [hepex], 2011.
[137] P. Adamson et al. Search for flavor-changing non-standard neutrino interactions by
MINOS. Phys. Rev., D88(7):072011, 2013.
[138] S. Bergmann, M. M. Guzzo, P. C. de Holanda, P. I. Krastev, and H. Nunokawa.
Status of the solution to the solar neutrino problem based on nonstandard neutrino
interactions. Phys. Rev., D62:073001, 2000.
[139] Zurab Berezhiani, R. S. Raghavan, and Anna Rossi. Probing non-standard couplings
of neutrinos at the borexino detector. Nucl. Phys., B638:62–80, 2002.
[140] A. Bolaños, O. G. Miranda, A. Palazzo, M. A. Tórtola, and J. W. F. Valle. Probing
non-standard neutrino-electron interactions with solar and reactor neutrinos. Phys.
Rev., D79:113012, 2009.
[141] A. Palazzo and J. W. F. Valle. Confusing non-zero θ13 with non-standard interactions in the solar neutrino sector. Phys. Rev., D80:091301, 2009.
[142] C. R. Das and Joao Pulido. Improving LMA predictions with non-standard interactions: Neutrino decay in solar matter? Phys. Rev., D83:053009, 2011.
[143] Antonio Palazzo. Hint of non-standard dynamics in solar neutrino conversion. Phys.
Rev., D83:101701, 2011.
[144] E.A. Garces, O.G. Miranda, M.A. Tórtola, and J. W. F. Valle. Low-energy neutrinoelectron scattering as a Standard Model probe: the potential of LENA as case study.
Phys.Rev., D85:073006, 2012.
[145] Sanjib Kumar Agarwalla, Francesco Lombardi, and Tatsu Takeuchi. Constraining
Non-Standard Interactions of the Neutrino with Borexino. JHEP, 12:079, 2012.
182
Bibliography
[146] M. C. Gonzalez-Garcia and Michele Maltoni. Determination of matter potential
from global analysis of neutrino oscillation data. JHEP, 09:152, 2013.
[147] Tzee-Ke Kuo and James T. Pantaleone. Three Neutrino Oscillations and the Solar
Neutrino Experiments. Phys. Rev., D35:3432, 1987.
[148] M. Deniz et al. Constraints on Non-Standard Neutrino Interactions and Unparticle Physics with Neutrino-Electron Scattering at the Kuo-Sheng Nuclear Power
Reactor. Phys. Rev., D82:033004, 2010.
[149] D. V. Forero and M. M. Guzzo. Constraining nonstandard neutrino interactions
with electrons. Phys. Rev., D84:013002, 2011.
[150] Stefan Antusch, Jochen P. Baumann, and Enrique Fernandez-Martinez. NonStandard Neutrino Interactions with Matter from Physics Beyond the Standard
Model. Nucl. Phys., B810:369–388, 2009.
[151] Zurab Berezhiani and Anna Rossi. Limits on the non-standard interactions of neutrinos from e+ e- colliders. Phys. Lett., B535:207–218, 2002.
[152] M. Deniz et al. Measurement of Neutrino-Electron Scattering Cross-Section with a
CsI(Tl) Scintillating Crystal Array at the Kuo-Sheng Nuclear Power Reactor. Phys.
Rev., D81:072001, 2010.
[153] Z. Daraktchieva et al. Limits on the neutrino magnetic moment from the MUNU
experiment. Phys. Lett., B564:190–198, 2003.
[154] A. I. Derbin et al. Experiment on anti-neutrino scattering by electrons at a reactor
of the Rovno nuclear power plant. JETP Lett., 57:768–772, 1993.
[155] G. S. Vidyakin, V. N. Vyrodov, I. I. Gurevich, Yu. V. Kozlov, V. P. Martemyanov,
S. V. Sukhotin, V. G. Tarasenkov, E. V. Turbin, and S. Kh. Khakhimov. Limitations
on the magnetic moment and charge radius of the electron-anti-neutrino. JETP
Lett., 55:206–210, 1992. [Pisma Zh. Eksp. Teor. Fiz.55,212(1992)].
[156] F. Reines, H. S. Gurr, and H. W. Sobel. Detection of anti-electron-neutrino e
scattering. Phys. Rev. Lett., 37:315–318, 1976.
[157] L. B. Auerbach et al. Measurement of electron-neutrino electron elastic scattering.
Phys. Rev., D63:112001, 2001.
Bibliography
183
[158] R. C. Allen et al. Study of electron-neutrino electron elastic scattering at LAMPF.
Phys. Rev., D47:11–28, 1993.
[159] P. Vilain et al. Precision measurement of electroweak parameters from the scattering
of muon-neutrinos on electrons. Phys. Lett., B335:246–252, 1994.
[160] J. V. Allaby et al. A Precise Determination of the Electroweak Mixing Angle from
Semileptonic Neutrino Scattering. Z. Phys., C36:611, 1987.
[161] A. Blondel, P. Bockmann, .H. Burkhardt, F. Dydak, A.L. Grant, et al. Electroweak Parameters From a High Statistics Neutrino Nucleon Scattering Experiment. Z.Phys., C45:361–379, 1990.
[162] G. P. Zeller et al. A Precise determination of electroweak parameters in neutrino nucleon scattering. Phys. Rev. Lett., 88:091802, 2002. [Erratum: Phys. Rev.
Lett.90,239902(2003)].
[163] Richard D. Ball et al. Precision determination of electroweak parameters and the
strange content of the proton from neutrino deep-inelastic scattering. Nucl.Phys.,
B823:195–233, 2009.
[164] W. Bentz, I. C. Cloet, J. T. Londergan, and A. W. Thomas. Reassessment of the
NuTeV determination of the Weinberg angle. Phys. Lett., B693:462–466, 2010.
[165] M. Altmann et al. Complete results for five years of GNO solar neutrino observations. Phys. Lett., B616:174–190, 2005.
[166] Florian Kaether. Datenanalyse der Sonnenneutrinoexperiments Gallex. PhD thesis,
Heidelberg U., 2007.
[167] S. Fukuda et al. Determination of solar neutrino oscillation parameters using 1496
days of Super-Kamiokande-I data. Phys. Lett., B539:179–187, 2002.
[168] S. N. Ahmed et al. Measurement of the total active B-8 solar neutrino flux at the
Sudbury Neutrino Observatory with enhanced neutral current sensitivity. Phys.
Rev. Lett., 92:181301, 2004.
[169] B. Aharmim et al. Electron energy spectra, fluxes, and day-night asymmetries of B8 solar neutrinos from the 391-day salt phase SNO data set. Phys. Rev., C72:055502,
2005.
184
Bibliography
[170] C. Arpesella et al. Direct Measurement of the Be-7 Solar Neutrino Flux with 192
Days of Borexino Data. Phys.Rev.Lett., 101:091302, 2008.
[171] C. Galbiati et al. New results on solar neutrino fluxes from 192 days of Borexino
data. J. Phys. Conf. Ser., 136:022001, 2008.
[172] G. Bellini et al. Measurement of the solar 8 B neutrino rate with a liquid scintillator target and 3 MeV energy threshold in the Borexino detector. Phys. Rev.,
D82:033006, 2010.
[173] L. Wolfenstein. Neutrino oscillations in matter. Phys. Rev., D17:2369, 1978.
[174] S. P. Mikheev and A. Yu. Smirnov. Resonance enhancement of oscillations in matter
and solar neutrino spectroscopy. Sov. J. Nucl. Phys., 42:913–917, 1985.
[175] A. Schaefer and S. E. Koonin. Influence of Density Fluctuations on Solar Neutrino
Conversion. Phys. Lett., B185:417–420, 1987.
[176] P. I. Krastev and A. Yu. Smirnov. Density perturbations and resonant conversion
of neutrinos. Mod. Phys. Lett., A6:1001–1009, 1991.
[177] F. N. Loreti and A. B. Balantekin. Neutrino oscillations in noisy media. Phys. Rev.,
D50:4762–4770, 1994.
[178] H. Nunokawa et al. The effect of random matter density perturbations on the MSW
solution to the solar neutrino problem. Nucl. Phys., B472:495–517, 1996.
[179] C. P. Burgess et al. Resonant origin for density fluctuations deep within the sun:
helioseismology and magneto-gravity waves. Mon. Not. Roy. Astron. Soc., 348:609,
2004.
[180] C. Burgess et al. Large mixing angle oscillations as a probe of the deep solar interior.
Astrophys. J., 588:L65, 2003.
[181] C. P. Burgess et al. Cornering solar radiative-zone fluctuations with KamLAND
and SNO salt. JCAP, 0401:007, 2004.
[182] G. L. Fogli, E. Lisi, A. Marrone, D. Montanino, and A. Palazzo. Probing nonstandard decoherence effects with solar and KamLAND neutrinos. Phys. Rev.,
D76:033006, 2007.
Bibliography
185
[183] J. Schechter and J. W. F. Valle. Majorana neutrinos and magnetic fields. Phys.
Rev., D24:1883, 1981. Err. D25, 283 (1982).
[184] E. Kh. Akhmedov. Resonant amplification of neutrino spin rotation in matter and
the solar-neutrino problem. Phys. Lett., B213:64–68, 1988.
[185] O. G. Miranda et al. The simplest resonant spin-flavour solution to the solar neutrino
problem. Nucl. Phys., B595:360–380, 2001.
[186] O. G. Miranda, T. I. Rashba, A. I. Rez, and J. W. F. Valle. Constraining the
neutrino magnetic moment with anti- neutrinos from the sun. Phys. Rev. Lett.,
93:051304, 2004.
[187] Alexander Friedland, Cecilia Lunardini, and Carlos Peña-Garay. Solar neutrinos as
probes of neutrino - matter interactions. Phys. Lett., B594:347, 2004.
[188] A. Bandyopadhyay et al. Physics at a future Neutrino Factory and super-beam
facility. Rept.Prog.Phys., 72:106201, 2009.
[189] Hiroshi Nunokawa, Stephen J. Parke, and Jose W.F. Valle. CP Violation and
Neutrino Oscillations. Prog.Part.Nucl.Phys., 60:338–402, 2008.
[190] T. Yanagida. KEK lectures. Ed. O. Sawada and A. Sugamoto (KEK, 1979), 1979.
[191] Rabindra N. Mohapatra and Goran Senjanovic. Neutrino mass and spontaneous
parity nonconservation. Phys. Rev. Lett., 44:91, 1980.
[192] F. Bazzocchi, D.G. Cerdeno, C. Munoz, and J. W. F. Valle. Calculable inverseseesaw neutrino masses in supersymmetry. Phys.Rev., D81:051701, 2010.
[193] Carlos Peña-Garay and Aldo Serenelli. Solar neutrinos and the solar composition
problem. ArXiv, 0811.2424 [astro-ph], 2008.
[194] J. Dorenbosch et al. Experimental Verification of the Universality of νe and νµ
Coupling to the Neutral Weak Current. Phys. Lett., B180:303, 1986.
[195] Andre de Gouvea and James Jenkins. What can we learn from neutrino electron
scattering? Phys. Rev., D74:033004, 2006.
[196] J. Barranco and others. Constraining non-standard interactions in nu/e e or antinu/e e scattering. Phys. Rev., D73:113001, 2006.
186
Bibliography
[197] Carla Biggio, Mattias Blennow, and Enrique Fernandez-Martinez. Loop bounds on
non-standard neutrino interactions. JHEP, 03:139, 2009.
[198] Stephen J. Parke. Nonadiabatic Level Crossing in Resonant Neutrino Oscillations.
Phys.Rev.Lett., 57:1275–1278, 1986.
[199] G.L. Fogli, E. Lisi, A. Marrone, D. Montanino, A. Palazzo, et al. Solar neutrino
oscillation parameters after first KamLAND results. Phys.Rev., D67:073002, 2003.
[200] G. L. Fogli, E. Lisi, A. Marrone, D. Montanino, and A. Palazzo. Getting the most
from the statistical analysis of solar neutrino oscillations. Phys. Rev., D66:053010,
2002.
[201] Yasuo Takeuchi. Low-energy neutrino observation at Super-Kamiokande-III. J.
Phys. Conf. Ser., 120:052008, 2008.
[202] Joshua R. Klein. Results and Prospects for SNO. Proceedings, 24th International
Conference on Neutrino physics and astrophysics (Neutrino 2010). Nucl. Phys.
Proc. Suppl., 229-232:68–73, 2012.
[203] Thomas Schwetz, M.A. Tórtola, and Jose W.F. Valle. Three-flavour neutrino oscillation update. New J.Phys., 10:113011, 2008.
[204] John N. Bahcall, K. Kubodera, and S. Nozawa. Neutral Current Reactions of Solar
and Supernova Neutrinos on Deuterium. Phys. Rev., D38:1030, 1988.
[205] K. Nakamura. Solar neutrino results, KamLAND and prospects. AIP Conf. Proc.,
721:12–19, 2004.
[206] K. Abe et al. Letter of Intent: The Hyper-Kamiokande Experiment — Detector
Design and Physics Potential —. ArXiv, 1109.3262 [hep-ex], 2011.
[207] K. Abe et al. A Long Baseline Neutrino Oscillation Experiment Using J-PARC
Neutrino Beam and Hyper-Kamiokande. ArXiv, 1412.4673, 2014.
[208] S. Davidson, S. Forte, P. Gambino, N. Rius, and A. Strumia. Old and new physics
interpretations of the NuTeV anomaly. JHEP, 0202:037, 2002.
[209] C.H. Llewellyn Smith. On the Determination of sin2 θw in Semileptonic Neutrino
Interactions. Nucl.Phys., B228:205, 1983.
Bibliography
187
[210] E.A. Paschos and L. Wolfenstein. Tests for neutral currents in neutrino reactions.
Phys.Rev., D7:91–95, 1973.
[211] K. Nakamura et al. Review of particle physics. J.Phys., G37:075021, 2010.
[212] Kevin Scott McFarland and Sven-Olaf Moch. Conventional physics explanations
for the NuTeV sin2 θW . Electroweak precision data and the Higgs mass. Proceedings, Workshop, Zeuthen, Germany, February 28-March 1, 2003 . ArXiv, hepph/0306052:61–83, 2003.
[213] M. Maltoni, T. Schwetz, M.A. Tórtola, and J.W.F. Valle. Status of global fits to
neutrino oscillations. New J.Phys., 6:122, 2004.
[214] M. C. Gonzalez-Garcia and Michele Maltoni. Phenomenology with Massive Neutrinos. Phys. Rept., 460:1–129, 2008.
[215] Gaku Mitsuka. Limit on non-standard interactions from the atmospheric neutrino
data. PoS, NUFACT08:059, 2008.
[216] A. M. Dziewonski and D. L. Anderson. Preliminary reference earth model. Phys.
Earth Planet. Interiors, 25:297–356, 1981.
[217] Eligio Lisi and Daniele Montanino. Earth regeneration effect in solar neutrino oscillations: An analytic approach. Phys. Rev., D56:1792–1803, 1997.
[218] D. G. Michael et al. Observation of muon neutrino disappearance with the MINOS
detectors and the NuMI neutrino beam. Phys. Rev. Lett., 97:191801, 2006.
[219] Mattias Blennow, Davide Meloni, Tommy Ohlsson, Francesco Terranova, and
Mattias Westerberg. Non-standard interactions using the OPERA experiment.
Eur.Phys.J., C56:529–536, 2008.
[220] R. B. Patterson. Oscillation measurements with the NuMI beam. Nucl. Part. Phys.
Proc., 265-266:153–158, 2015.
[221] C. Adams et al. The Long-Baseline Neutrino Experiment: Exploring Fundamental
Symmetries of the Universe. ArXiv, 1307.7335 [hep-ex].
[222] J. M. Conrad. Neutrino scattering on Glass: NuSOnG. AIP Conf. Proc., 981:243–
246, 2008.
188
Bibliography
[223] T. Adams et al. Terascale Physics Opportunities at a High Statistics, High Energy
Neutrino Scattering Experiment: NuSOnG. Int.J.Mod.Phys., A24:671–717, 2009.
[224] Evgeny K. Akhmedov, Manfred Lindner, Erhard Schnapka, and J.W.F. Valle. Leftright symmetry breaking in NJL approach. Phys.Lett., B368:270–280, 1996.
[225] Evgeny K. Akhmedov, Manfred Lindner, Erhard Schnapka, and J.W.F. Valle. Dynamical left-right symmetry breaking. Phys.Rev., D53:2752–2780, 1996.
[226] Sofiane M. Boucenna, Stefano Morisi, and José W.F. Valle. The low-scale approach
to neutrino masses. Adv.High Energy Phys., 2014:831598, 2014.
[227] P.S. Bhupal Dev and R.N. Mohapatra. TeV Scale Inverse Seesaw in SO(10) and
Leptonic Non-Unitarity Effects. Phys.Rev., D81:013001, 2010.
[228] Chang-Hun Lee, P.S. Bhupal Dev, and R.N. Mohapatra. Natural TeV-scale left-right
seesaw mechanism for neutrinos and experimental tests. Phys.Rev., D88(9):093010,
2013.
[229] M. Dittmar, A. Santamaria, M. C. Gonzalez-Garcia, and J. W. F. Valle. Production
mechanisms and signatures of isosinglet neutral heavy leptons in Z0 decays. Nucl.
Phys., B332:1, 1990.
[230] M.Z. Akrawy et al. Limits on neutral heavy lepton production from Z0 decay.
Phys.Lett., B247:448–457, 1990.
[231] P. Abreu et al. Search for neutral heavy leptons produced in Z decays. Z.Phys.,
C74:57–71, 1997.
[232] A. Aguilar et al. Evidence for neutrino oscillations from the observation of anti-nu/e
appearance in a anti-nu/mu beam. Phys. Rev., D64:112007, 2001.
[233] A.A. Aguilar-Arevalo et al. Improved Search for ν̄µ → ν̄e Oscillations in the MiniBooNE Experiment. Phys.Rev.Lett., 110:161801, 2013.
[234] Takehiko Asaka, Steve Blanchet, and Mikhail Shaposhnikov. The nuMSM, dark
matter and neutrino masses. Phys.Lett., B631:151–156, 2005.
[235] Robert E. Shrock. General Theory of Weak Leptonic and Semileptonic Decays. 1.
Leptonic Pseudoscalar Meson Decays, with Associated Tests For, and Bounds on,
Neutrino Masses and Lepton Mixing. Phys.Rev., D24:1232, 1981.
Bibliography
189
[236] Michael Gronau, Chung Ngoc Leung, and Jonathan L. Rosner. Extending Limits
on Neutral Heavy Leptons. Phys.Rev., D29:2539, 1984.
[237] J. Schechter and J.W.F. Valle. Neutrinoless Double beta Decay in SU(2) x U(1)
Theories. Phys.Rev., D25:2951, 1982.
[238] Martin Hirsch. Phenomenology of neutrinoless double beta decay. Nucl. Phys. Proc.
Suppl., 221:119–124, 2011.
[239] Werner Rodejohann. Neutrino-less Double Beta Decay and Particle Physics.
Int.J.Mod.Phys., E20:1833–1930, 2011.
[240] J. Bernabeu et al. Lepton flavor nonconservation at high-energies in a superstring
inspired standard model. Phys. Lett., B187:303, 1987.
[241] Hans Hettmansperger, Manfred Lindner, and Werner Rodejohann. Phenomenological Consequences of sub-leading Terms in See-Saw Formulas. JHEP, 1104:123,
2011.
[242] G. C. Branco, M. N. Rebelo, and J. W. F. Valle. Leptonic CP violation with massless
neutrinos. Phys. Lett., B225:385, 1989.
[243] N. Rius and J. W. F. Valle. Leptonic CP violating asymmetries in Z0 decays. Phys.
Lett., B246:249–255, 1990.
[244] Enrico Nardi, Esteban Roulet, and Daniele Tommasini. Limits on neutrino mixing
with new heavy particles. Phys.Lett., B327:319–326, 1994.
[245] Paul Langacker and David London. Mixing Between Ordinary and Exotic Fermions.
Phys.Rev., D38:886, 1988.
[246] M.C. Gonzalez-Garcia, A. Santamaria, and J.W.F. Valle. Isosinglet Neutral Heavy
Lepton Production in Z Decays and Neutrino Mass. Nucl.Phys., B342:108–126,
1990.
[247] A. Abada, D. Das, A.M. Teixeira, A. Vicente, and C. Weiland. Tree-level lepton
universality violation in the presence of sterile neutrinos: impact for RK and Rπ .
JHEP, 1302:048, 2013.
[248] A. Abada, A.M. Teixeira, A. Vicente, and C. Weiland. Sterile neutrinos in leptonic
and semileptonic decays. JHEP, 1402:091, 2014.
190
Bibliography
[249] Anupama Atre, Tao Han, Silvia Pascoli, and Bin Zhang. The Search for Heavy
Majorana Neutrinos. JHEP, 0905:030, 2009.
[250] A. Abada, V. De Romeri, and A.M. Teixeira. Effect of steriles states on lepton
magnetic moments and neutrinoless double beta decay. JHEP, 1409:074, 2014.
[251] Stefan Antusch and Oliver Fischer. Non-unitarity of the leptonic mixing matrix:
Present bounds and future sensitivities. JHEP, 1410:94, 2014.
[252] G. Czapek et al. Branching ratio for the rare pion decay into positron and neutrino.
Phys. Rev. Lett., 70:17–20, 1993.
[253] Bernard Aubert et al. Measurements of Charged Current Lepton Universality and
—V(us)— using Tau Lepton Decays to e- nu(e)-bar nu(tau), mu-bar nu(mu)-bar
nu(tau), pi- nu(tau) and K- nu(tau). Phys.Rev.Lett., 105:051602, 2010.
[254] Martin Freund. Analytic approximations for three neutrino oscillation parameters
and probabilities in matter. Phys. Rev., D64:053003, 2001.
[255] Evgeny K. Akhmedov, Robert Johansson, Manfred Lindner, Tommy Ohlsson, and
Thomas Schwetz. Series expansions for three flavor neutrino oscillation probabilities
in matter. JHEP, 04:078, 2004.
[256] C. Giunti, M. Laveder, Y.F. Li, Q.Y. Liu, and H.W. Long. Update of Short-Baseline
Electron Neutrino and Antineutrino Disappearance. Phys.Rev., D86:113014, 2012.
[257] C. Giunti, M. Laveder, Y.F. Li, and H.W. Long. Pragmatic View of Short-Baseline
Neutrino Oscillations. Phys.Rev., D88:073008, 2013.
[258] G. Mention et al. The Reactor Antineutrino Anomaly. Phys. Rev., D83:073006,
2011.
[259] Tzee-Ke Kuo and James T. Pantaleone. Neutrino oscillations in matter. Rev. Mod.
Phys., 61:937, 1989.
[260] P. Astier et al. Search for νµ → νe oscillations in the NOMAD experiment. Phys.
Lett., B570:19–31, 2003.
[261] Marco Cirelli, Guido Marandella, Alessandro Strumia, and Francesco Vissani. Probing oscillations into sterile neutrinos with cosmology, astrophysics and experiments.
Nucl. Phys., B708:215–267, 2005.
Bibliography
191
[262] ATLAS Collaboration. Search for Majorana neutrino production in pp collisions at
sqrt(s)=7 TeV in dimuon final states with the ATLAS detector. ATLAS-CONF2012-139, ATLAS-COM-CONF-2012-165, 2012.
[263] Vardan Khachatryan et al. Search for heavy Majorana neutrinos in µ± µ± + jets
√
events in proton-proton collisions at s = 8 TeV. Phys. Lett., B748:144–166, 2015.
[264] Roel Aaij et al. Search for Majorana neutrinos in B − → π + µ− µ− decays.
Phys.Rev.Lett., 112(13):131802, 2014.
[265] D.I. Britton, S. Ahmad, D.A. Bryman, R.A. Burnbam, E.T.H. Clifford, et al. Measurement of the π + → e+ neutrino branching ratio. Phys.Rev.Lett., 68:3000–3003,
1992.
[266] D.I. Britton, S. Ahmad, D.A. Bryman, R.A. Burnham, E.T.H. Clifford, et al. Improved search for massive neutrinos in π + → e+ neutrino decay. Phys.Rev., D46:885–
887, 1992.
[267] G. Bernardi, G. Carugno, J. Chauveau, F. Dicarlo, M. Dris, et al. Further limits
on heavy neutrino couplings. Phys.Lett., B203:332, 1988.
[268] J. Badier et al. Mass and Lifetime Limits on New Longlived Particles in 300GeV/cπ − Interactions. Z.Phys., C31:21, 1986.
[269] F. Bergsma et al. A Search for Decays of Heavy Neutrinos in the Mass Range
0.5-GeV to 2.8-GeV. Phys.Lett., B166:473, 1986.
[270] D. Liventsev et al. Search for heavy neutrinos at Belle. Phys.Rev., D87(7):071102,
2013.
[271] O. Adriani et al. Search for isosinglet neutral heavy leptons in Z0 decays. Phys.Lett.,
B295:371–382, 1992.
[272] P. Achard et al. Search for heavy isosinglet neutrino in e+ e− annihilation at LEP.
Phys.Lett., B517:67–74, 2001.
√
[273] J. Klinger. Search for heavy majorana neutrinos in pp collisions at s = 8 tev with
the atlas detector. University of Manchester, Ph D thesis, B388:81–112, 2014.
[274] Frank F. Deppisch, P. S. Bhupal Dev, and Apostolos Pilaftsis. Neutrinos and Collider Physics. New J. Phys., 17(7):075019, 2015.
192
Bibliography
[275] Strahinja Lukić. Physics highlights at ILC and CLIC. 12th International SchoolSeminar on The Actual Problems of Microworld Physics Gomel, Belarus, July 22August 2, 2013 . ArXiv, 1406.4313, 2014.
[276] Frank Zimmermann et al. FCC-ee Overview. Proceedings, 55th ICFA Advanced
Beam Dynamics Workshop on High Luminosity Circular e+e- Colliders – Higgs
Factory (HF2014). THP3H1, 2015.
[277] A. Abada, V. De Romeri, S. Monteil, J. Orloff, and A. M. Teixeira. Indirect searches
for sterile neutrinos at a high-luminosity Z-factory. JHEP, 04:051, 2015.
[278] R.S. Hayano, T. Taniguchi, T. Yamanaka, T. Tanimori, R. Enomoto, et al. Heavy
neutrino search using K(mu2) decay. Phys.Rev.Lett., 49:1305, 1982.
[279] Alexander Kusenko, Silvia Pascoli, and Dmitry Semikoz. New bounds on MeV
sterile neutrinos based on the accelerator and Super-Kamiokande results. JHEP,
0511:028, 2005.
[280] P. Vilain et al. Search for heavy isosinglet neutrinos. Phys.Lett., B343:453–458,
1995.
[281] E. Gallas et al. Search for neutral weakly interacting massive particles in the Fermilab Tevatron wide band neutrino beam. Phys.Rev., D52:6–14, 1995.
[282] Amanda M. Cooper-Sarkar et al. Search for Heavy Neutrino Decays in the BEBC
Beam Dump Experiment. Phys.Lett., B160:207, 1985.
[283] A. Vaitaitis et al. Search for neutral heavy leptons in a high-energy neutrino beam.
Phys.Rev.Lett., 83:4943–4946, 1999.
[284] A.V. Artamonov et al. Search for heavy neutrinos in K + → µ+ νH decays. Phys.Rev.,
D91(5):052001, 2015.
[285] P. Astier et al. Search for heavy neutrinos mixing with tau neutrinos. Phys.Lett.,
B506:27–38, 2001.
[286] J. Orloff, Alexandre N. Rozanov, and C. Santoni. Limits on the mixing of tau
neutrino to heavy neutrinos. Phys.Lett., B550:8–15, 2002.
[287] J.A. Aguilar-Saavedra, F. Deppisch, O. Kittel, and J. W. F. Valle. Flavour in heavy
neutrino searches at the LHC. Phys.Rev., D85:091301, 2012.
Bibliography
193
[288] S.P. Das, F.F. Deppisch, O. Kittel, and J. W. F. Valle. Heavy Neutrinos and
Lepton Flavour Violation in Left-Right Symmetric Models at the LHC. Phys.Rev.,
D86:055006, 2012.
[289] Frank F. Deppisch, Martin Hirsch, and Heinrich Pas. Neutrinoless Double Beta
Decay and Physics Beyond the Standard Model. J.Phys., G39:124007, 2012.
[290] Manimala Mitra, Goran Senjanovic, and Francesco Vissani. Neutrinoless Double
Beta Decay and Heavy Sterile Neutrinos. Nucl.Phys., B856:26–73, 2012.
[291] Frank F. Deppisch, Nishita Desai, and Jose W. F. Valle. Is charged lepton flavour
violation a high energy phenomenon? Phys.Rev., D89:051302(R), 2014.
[292] Chien-Yi Chen, P. S. Bhupal Dev, and R.N. Mohapatra. Probing Heavy-Light
Neutrino Mixing in Left-Right Seesaw Models at the LHC. Phys.Rev., D88:033014,
2013.
[293] Michael Wurm et al. The next-generation liquid-scintillator neutrino observatory
LENA. Astropart. Phys., 35:685–732, 2012.
[294] T. Patzak. LAGUNA-LBNO: Large apparatus studying grand unification and neutrino astrophysics and long baseline neutrino oscillations. J. Phys. Conf. Ser.,
375:042056, 2012.
[295] Irina Mocioiu and Warren Wright. Non-standard neutrino interactions in the
mu–tau sector. Nucl. Phys., B893:376–390, 2015.
[296] J. P. Yanez. Results from atmospheric neutrino oscillations with IceCube DeepCore.
AIP Conf. Proc., 1666:100002, 2015.
[297] F. J. Escrihuela, D. V. Forero, O. G. Miranda, M. Tórtola, and J. W. F. Valle. Constraining right-handed neutrinos. International Conference on High Energy Physics
2014 (ICHEP 2014) Valencia, Spain, July 2-9, 2014 . ArXiv, 1505.01097 [hep-ph],
2015.
[298] Double Chooz Experiment webpage.
ner/index.php.
http://www.physics.drexel.edu/ wkush-
[299] Xinheng Guo et al. A precision measurement of the neutrino mixing angle θ13 using
reactor antineutrinos at Daya Bay. ArXiv, 0701029 [hep-ex], 2007.
194
Bibliography
[300] F. P. An et al. A side-by-side comparison of Daya Bay antineutrino detectors. Nucl.
Instrum. Meth., A685:78–97, 2012.
[301] CERN Courier webpage. http://cerncourier.com/cws/article/cern/31840.
[302] Seon-Hee Seo. New Results from RENO. Proceedings of the 15th International
Workshop on Neutrino Telescopes (Neutel 2013). ArXiv, 1312.4111, 2014.
[303] A. Suzuki. Results from KamLAND reactor neutrino detection. Phys. Scripta,
T121:33–38, 2005.
[304] SNO Experiment webpage. http://www.sno.phy.queensu.ca/.
[305] Super-Kamiokande webpage. http://www-sk.icrr.u-tokyo.ac.jp/.
[306] G. Bellini, L. Ludhova, G. Ranucci, and F. L. Villante. Neutrino oscillations. Adv.
High Energy Phys., 2014:191960, 2014.
[307] J. V. Allaby et al. A Precise Determination of the Electroweak Mixing Angle from
Semileptonic Neutrino Scattering. Phys. Lett., B177:446, 1986.
[308] NuTeV Experiment webpage. http://www-e815.fnal.gov/.
[309] D. Geiregat et al. Calibration and performance of the CHARM-II detector. Nucl.
Instrum. Meth., A325:92–108, 1993.